Advances in Smart Grid Technology: Select Proceedings of PECCON 2019—Volume I [1st ed.] 9789811572449, 9789811572456

This book comprises the select proceedings of the International Conference on Power Engineering Computing and Control (P

311 60 28MB

English Pages XI, 537 [531] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Matter ....Pages i-xi
Front Matter ....Pages 1-1
Z-Source Inverter-Based DVR Using Wind Generator as a Source for Storage Unit (Durai Babu, P. Murugesan)....Pages 3-12
Predicting Wind Turbine Blade Fault Condition to Enhance Wind Energy Harvest Through Classification via Regression Classifier (A. Joshuva, M. Arjun, R. Murugavel, V. A. Shridhar, G. S. Sriram Gangadhar, S. S. Dhanush)....Pages 13-20
Satin Bower Bird Algorithm for Controller Parameter Optimization in an Autonomous AC Microgrid (Rekha P. Nair, P. Kanakasabapathy)....Pages 21-30
Modeling and Simulation of a DFIG-Based Wind Energy System (Madhuvanthani Rajendran, L. Ashok Kumar)....Pages 31-49
Development of Wind Energy Technologies and Their Impact on Environment: A Review (Manyamyuva Naga Satya Suryakiran, Waseemah Begum, R. S. Sudhakar, Sharad Kumar Tiwari)....Pages 51-62
Regulated Jordan—Elman Neural Network-Based Controller Model for Grid-Connected Wind Energy Conversion Systems (S. N. Deepa, N. Rajasingam)....Pages 63-75
BESS-Based Microgrid with Enhanced Power Control and Storage Management (Shanmugam Kalpana, Gogineni Pradeep)....Pages 77-92
Adaptive Reserve Estimation Technique for PV Systems Under Dynamic Operating Conditions (Pankaj Verma, Tarlochan Kaur, Raminder Kaur)....Pages 93-104
Assessing Small Cross Flow Wind Turbine for Urban Rooftop Power Generation (Seralathan Sivamani, R. Hemanth Prasanna, J. Arun, Mikhail Christopher, T. Micha Premkumar, P. Bharath Kumar et al.)....Pages 105-114
A Novel Power Electronic-Based Maximum Power Point Tracking Technique for Solar PV Applications (D. Ravi Kishore, T. Vijay Muni, K. S. Srikanth)....Pages 115-126
Analysis of an Enhanced Positive Output Super-Lift Luo Converter for Renewable Energy Applications (A. Paramasivam, K. B. Bhaskar, N. Madhanakkumar, C. Vanchinathan)....Pages 127-136
Power Quality Enhancement of DC Micro-grid Using DC Electric Spring (A. G. Anu, R. Hari Kumar, S. Ushakumari)....Pages 137-148
Mathematical Modelling of Embedded Switched-Inductor Z-Source Inverter for Photovoltaic Energy Conversion (T. Divya, R. Ramaprabha)....Pages 149-164
Hybrid Algorithms to Track Peak Power in Solar PV Array Under All Irradiation Conditions (R. Ramaprabha, S. Malathy)....Pages 165-177
Financial Analysis of Diesel and Solar Photovoltaic Water Pumping Systems (M. Pandikumar, R. Ramaprabha)....Pages 179-188
Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode Controller in PMSG-Based Wind Turbines (M. Rajvikram, J. Vishnupriyan)....Pages 189-201
Sustainable Energy Technologies: Energy Resources for Portable Electronics—A Mini Review (M. Malini, P. Sriramalakshmi, M. Sujatha)....Pages 203-215
Front Matter ....Pages 217-217
Simulation and Analysis of a Voltage Control Strategy for Single-Stage AC-AC Converter (A. Jamna, R. Sujatha, V. Jamuna)....Pages 219-230
Implementing PV Energized SRSPM-Based Single-Phase Inverter for Induction Motor (P. Abirami, M. Pushpavalli, P. Sivagami)....Pages 231-246
Enhancement of Voltage Transfer Ratio with Matrix Converter Based on Modulation Duty Cycle Matrix Approach Using Optimum AV Method (N. Krishna Kumari, D. Ravi Kumar, E. Shiva Prasad)....Pages 247-262
Design and Implementation of Feedback Controller for Nonisolated Switching DC-DC Buck Converter Operating in Continuous Current Conduction Mode (Manthan Mangesh Borage, Dipen M. Vachhani, Rajesh Arya)....Pages 263-276
Design and Analysis of Improved Indirect Matrix Converter Supplying Power to Rotor of DFIG for Bi-directional Power Flow (N. Lavanya, P. N. H. Phanindra Kumar)....Pages 277-287
Dynamic Analysis of Fuel Cell-Fed Superlift Converter (R. Hounandan, V. Chamundeeswari, M. S. Kumaran, J. Deepa)....Pages 289-305
DSPACE1103 Controller for PWM Control of Power Electronic Converters (R. Amalrajan, R. Gunabalan, Nilanjan Tewari)....Pages 307-322
Performance and Reliability Analysis of 13-Level Asymmetrical Inverter with Reduced Devices (Abeera Dutt Roy, Chandrahasan Umayal)....Pages 323-335
Lighting Design for a Campus Hostel Building Using Single-Stage Efficient Buck LED Driver (R. Srimathi, Kishore Eswaran, S. Hemamalini, V. Kamatchi Kannan)....Pages 337-351
Front Matter ....Pages 351-351
A Novel Approach for Hobby Class Remotely Operated Vehicle (T. M. Thamizh Thentral, T. V. Abhinav Viswanaath, S. Senthilnathan, R. Bhargav)....Pages 355-368
Simulation and Analysis of Interleaved Buck DC-DC Converter for EV Charging (K. Murugappan, R. Seyezhai, G. Kishor Sabarish, N. Kaashyap, J. Jason Ranjit)....Pages 369-380
Design and Real-Time Implementation of Economical Solar Car for Commercial Applications (Nabeel Najeeb Kassim, Sohail Akthar Khan, Kumar Devasish, A. Chitra, Ryan Sujith)....Pages 381-395
Electric Vehicles Integration with Renewable Energy Sources and Smart Grids (G. Sree Lakshmi, Rubanenko Olena, G. Divya, I. Hunko)....Pages 397-411
Front Matter ....Pages 413-413
FOPD Controller Using Bee Colony Optimized Reduced Order FOFOPDT Model of Three Interacting Tank Process (U. Sabura Banu, Abdul Wahid Nasir, I. Mohamed Shiek Mothi)....Pages 415-434
Hardware Security for DSP Circuits Using Key-Based Time-Dependent Obfuscation (R. Uma Rani, D. Jayanthi, Sudharsan Jayabalan, Arun Vignesh)....Pages 435-444
Assessing the Performance of Response Surface Methodology on the Transesterification Process of Moringa Seed Bio-oil (V. Hariram, H. K. Ganesh, Patan Ahmad, Mallela Krishna Kumar, Diviti Somasekhar, S. Seralathan et al.)....Pages 445-458
Optimizing the Transesterification Process of Hemp Seed Bio-oil Using Artificial Neural Network (V. Hariram, P. M. Bharadwaj, A. Viswaksen, C. H. Surya, D. Ruthvin Maheej, S. Seralathan et al.)....Pages 459-471
Development and Performance Evaluation of Micromagnesium Oxide Filled Silicone Rubber for High Voltage Insulation (Vinayak. V. Rao, Murthy K. Ramakrishana, Pradipkumar Dixit, R. Harikrishna, M. Anand, R. T. Prajwal Kumar et al.)....Pages 473-484
Performance Evaluation of Naturally Aged Silicone Rubber Polymeric Insulating Specimen with Nano TiO2 Filled Epoxy Coating (K. Ramakrishana Murthy, Vinayak. V. Rao, Pradipkumar Dixit, R. Harikrishna, Tarun Anand, Chandrakala et al.)....Pages 485-496
Ontology-Based Crime Investigation Process (P. Venkata Srimukh, S. Shridevi)....Pages 497-509
Variation-Tolerant In-Memory Digital Computations Using SRAM (D. Gracin, Kichchannagari Omkar, V. Ravi, T. Vigneswaran)....Pages 511-522
Defect Detection and Defect-Tolerant Design of a Multi-port SRAM (R. Sushmitha, O. R. Nisha, V. Ravi, T. Vigneswaran)....Pages 523-537
Recommend Papers

Advances in Smart Grid Technology: Select Proceedings of PECCON 2019—Volume I [1st ed.]
 9789811572449, 9789811572456

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Lecture Notes in Electrical Engineering 687

Pierluigi Siano K. Jamuna   Editors

Advances in Smart Grid Technology Select Proceedings of PECCON 2019— Volume I

Lecture Notes in Electrical Engineering Volume 687

Series Editors Leopoldo Angrisani, Department of Electrical and Information Technologies Engineering, University of Napoli Federico II, Naples, Italy Marco Arteaga, Departament de Control y Robótica, Universidad Nacional Autónoma de México, Coyoacán, Mexico Bijaya Ketan Panigrahi, Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi, India Samarjit Chakraborty, Fakultät für Elektrotechnik und Informationstechnik, TU München, Munich, Germany Jiming Chen, Zhejiang University, Hangzhou, Zhejiang, China Shanben Chen, Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China Tan Kay Chen, Department of Electrical and Computer Engineering, National University of Singapore, Singapore, Singapore Rüdiger Dillmann, Humanoids and Intelligent Systems Laboratory, Karlsruhe Institute for Technology, Karlsruhe, Germany Haibin Duan, Beijing University of Aeronautics and Astronautics, Beijing, China Gianluigi Ferrari, Università di Parma, Parma, Italy Manuel Ferre, Centre for Automation and Robotics CAR (UPM-CSIC), Universidad Politécnica de Madrid, Madrid, Spain Sandra Hirche, Department of Electrical Engineering and Information Science, Technische Universität München, Munich, Germany Faryar Jabbari, Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA, USA Limin Jia, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Alaa Khamis, German University in Egypt El Tagamoa El Khames, New Cairo City, Egypt Torsten Kroeger, Stanford University, Stanford, CA, USA Qilian Liang, Department of Electrical Engineering, University of Texas at Arlington, Arlington, TX, USA Ferran Martín, Departament d’Enginyeria Electrònica, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain Tan Cher Ming, College of Engineering, Nanyang Technological University, Singapore, Singapore Wolfgang Minker, Institute of Information Technology, University of Ulm, Ulm, Germany Pradeep Misra, Department of Electrical Engineering, Wright State University, Dayton, OH, USA Sebastian Möller, Quality and Usability Laboratory, TU Berlin, Berlin, Germany Subhas Mukhopadhyay, School of Engineering & Advanced Technology, Massey University, Palmerston North, Manawatu-Wanganui, New Zealand Cun-Zheng Ning, Electrical Engineering, Arizona State University, Tempe, AZ, USA Toyoaki Nishida, Graduate School of Informatics, Kyoto University, Kyoto, Japan Federica Pascucci, Dipartimento di Ingegneria, Università degli Studi “Roma Tre”, Rome, Italy Yong Qin, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China Gan Woon Seng, School of Electrical & Electronic Engineering, Nanyang Technological University, Singapore, Singapore Joachim Speidel, Institute of Telecommunications, Universität Stuttgart, Stuttgart, Germany Germano Veiga, Campus da FEUP, INESC Porto, Porto, Portugal Haitao Wu, Academy of Opto-electronics, Chinese Academy of Sciences, Beijing, China Junjie James Zhang, Charlotte, NC, USA

The book series Lecture Notes in Electrical Engineering (LNEE) publishes the latest developments in Electrical Engineering—quickly, informally and in high quality. While original research reported in proceedings and monographs has traditionally formed the core of LNEE, we also encourage authors to submit books devoted to supporting student education and professional training in the various fields and applications areas of electrical engineering. The series cover classical and emerging topics concerning: • • • • • • • • • • • •

Communication Engineering, Information Theory and Networks Electronics Engineering and Microelectronics Signal, Image and Speech Processing Wireless and Mobile Communication Circuits and Systems Energy Systems, Power Electronics and Electrical Machines Electro-optical Engineering Instrumentation Engineering Avionics Engineering Control Systems Internet-of-Things and Cybersecurity Biomedical Devices, MEMS and NEMS

For general information about this book series, comments or suggestions, please contact [email protected]. To submit a proposal or request further information, please contact the Publishing Editor in your country: China Jasmine Dou, Associate Editor ([email protected]) India, Japan, Rest of Asia Swati Meherishi, Executive Editor ([email protected]) Southeast Asia, Australia, New Zealand Ramesh Nath Premnath, Editor ([email protected]) USA, Canada: Michael Luby, Senior Editor ([email protected]) All other Countries: Leontina Di Cecco, Senior Editor ([email protected]) ** Indexing: The books of this series are submitted to ISI Proceedings, EI-Compendex, SCOPUS, MetaPress, Web of Science and Springerlink **

More information about this series at http://www.springer.com/series/7818

Pierluigi Siano K. Jamuna •

Editors

Advances in Smart Grid Technology Select Proceedings of PECCON 2019—Volume I

123

Editors Pierluigi Siano Department of Management and Innovation Systems University of Salerno Fisciano, Italy

K. Jamuna School of Electrical Engineering Vellore Institute of Technology Chennai, India

ISSN 1876-1100 ISSN 1876-1119 (electronic) Lecture Notes in Electrical Engineering ISBN 978-981-15-7244-9 ISBN 978-981-15-7245-6 (eBook) https://doi.org/10.1007/978-981-15-7245-6 © Springer Nature Singapore Pte Ltd. 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Power system is undergoing tremendous developments such as integration of more and more renewable energy resources, advancements in power electronic interfaces, energy efficiency measures, control and introduction of more intelligence in the grid. This book constitutes selected high-quality papers presented in the Second International Conference on Power Engineering, Computing and CONtrol, PECCON 2019, during 12 to 14 December 2019. The main theme of PECCON 2019 is “smart technology for smart grid”. The conference provided a forum for practicing professionals, academicians, research scholars and students to exchange their innovative ideas, inferences and knowledge gathered through rigorous experiments. This book volume consists of 39 chapters and is organized into four parts—Sustainable Energy Technologies, Power Electronics, Electric Vehicles and Allied Technologies. This book will be useful to the researchers, academicians, students and professionals working in the field as well as R&D organizations in the domain of electrical power and energy infrastructure. We, the editors of this book, are thankful to all the authors who have contributed to the paper submission. We also thank all the reviewers and technical committee members for providing their valuable comments on time and help towards the improvement of the quality of papers presented in the conference. We also thank members of the advisory board and session chairs for their valuable comments. Our special thanks to Series Editors, Lecture Notes in Electrical Engineering, Springer, for giving us the opportunity to publish this edited volume in the series. Chennai, India Fisciano, Italy

K. Jamuna Pierluigi Siano

v

Contents

Sustainable Energy Technologies Z-Source Inverter-Based DVR Using Wind Generator as a Source for Storage Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Durai Babu and P. Murugesan Predicting Wind Turbine Blade Fault Condition to Enhance Wind Energy Harvest Through Classification via Regression Classifier . . . . . A. Joshuva, M. Arjun, R. Murugavel, V. A. Shridhar, G. S. Sriram Gangadhar, and S. S. Dhanush Satin Bower Bird Algorithm for Controller Parameter Optimization in an Autonomous AC Microgrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rekha P. Nair and P. Kanakasabapathy Modeling and Simulation of a DFIG-Based Wind Energy System . . . . . Madhuvanthani Rajendran and L. Ashok Kumar Development of Wind Energy Technologies and Their Impact on Environment: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manyamyuva Naga Satya Suryakiran, Waseemah Begum, R. S. Sudhakar, and Sharad Kumar Tiwari

3

13

21 31

51

Regulated Jordan—Elman Neural Network-Based Controller Model for Grid-Connected Wind Energy Conversion Systems . . . . . . . . . . . . . S. N. Deepa and N. Rajasingam

63

BESS-Based Microgrid with Enhanced Power Control and Storage Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shanmugam Kalpana and Gogineni Pradeep

77

Adaptive Reserve Estimation Technique for PV Systems Under Dynamic Operating Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pankaj Verma, Tarlochan Kaur, and Raminder Kaur

93

vii

viii

Contents

Assessing Small Cross Flow Wind Turbine for Urban Rooftop Power Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Seralathan Sivamani, R. Hemanth Prasanna, J. Arun, Mikhail Christopher, T. Micha Premkumar, P. Bharath Kumar, Yeswanth Yadav, and V. Hariram A Novel Power Electronic-Based Maximum Power Point Tracking Technique for Solar PV Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 D. Ravi Kishore, T. Vijay Muni, and K. S. Srikanth Analysis of an Enhanced Positive Output Super-Lift Luo Converter for Renewable Energy Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 A. Paramasivam, K. B. Bhaskar, N. Madhanakkumar, and C. Vanchinathan Power Quality Enhancement of DC Micro-grid Using DC Electric Spring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 A. G. Anu, R. Hari Kumar, and S. Ushakumari Mathematical Modelling of Embedded Switched-Inductor Z-Source Inverter for Photovoltaic Energy Conversion . . . . . . . . . . . . . . . . . . . . . 149 T. Divya and R. Ramaprabha Hybrid Algorithms to Track Peak Power in Solar PV Array Under All Irradiation Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 R. Ramaprabha and S. Malathy Financial Analysis of Diesel and Solar Photovoltaic Water Pumping Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 M. Pandikumar and R. Ramaprabha Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode Controller in PMSG-Based Wind Turbines . . . . . . . . . . . . . . . . . 189 M. Rajvikram and J. Vishnupriyan Sustainable Energy Technologies: Energy Resources for Portable Electronics—A Mini Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 M. Malini, P. Sriramalakshmi, and M. Sujatha Power Electronics Simulation and Analysis of a Voltage Control Strategy for Single-Stage AC-AC Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 A. Jamna, R. Sujatha, and V. Jamuna Implementing PV Energized SRSPM-Based Single-Phase Inverter for Induction Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 P. Abirami, M. Pushpavalli, and P. Sivagami

Contents

ix

Enhancement of Voltage Transfer Ratio with Matrix Converter Based on Modulation Duty Cycle Matrix Approach Using Optimum AV Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 N. Krishna Kumari, D. Ravi Kumar, and E. Shiva Prasad Design and Implementation of Feedback Controller for Nonisolated Switching DC-DC Buck Converter Operating in Continuous Current Conduction Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Manthan Mangesh Borage, Dipen M. Vachhani, and Rajesh Arya Design and Analysis of Improved Indirect Matrix Converter Supplying Power to Rotor of DFIG for Bi-directional Power Flow . . . . 277 N. Lavanya and P. N. H. Phanindra Kumar Dynamic Analysis of Fuel Cell-Fed Superlift Converter . . . . . . . . . . . . . 289 R. Hounandan, V. Chamundeeswari, M. S. Kumaran, and J. Deepa DSPACE1103 Controller for PWM Control of Power Electronic Converters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 R. Amalrajan, R. Gunabalan, and Nilanjan Tewari Performance and Reliability Analysis of 13-Level Asymmetrical Inverter with Reduced Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323 Abeera Dutt Roy and Chandrahasan Umayal Lighting Design for a Campus Hostel Building Using Single-Stage Efficient Buck LED Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 R. Srimathi, Kishore Eswaran, S. Hemamalini, and V. Kamatchi Kannan Electric Vehicles A Novel Approach for Hobby Class Remotely Operated Vehicle . . . . . . 355 T. M. Thamizh Thentral, T. V. Abhinav Viswanaath, S. Senthilnathan, and R. Bhargav Simulation and Analysis of Interleaved Buck DC-DC Converter for EV Charging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 K. Murugappan, R. Seyezhai, G. Kishor Sabarish, N. Kaashyap, and J. Jason Ranjit Design and Real-Time Implementation of Economical Solar Car for Commercial Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 Nabeel Najeeb Kassim, Sohail Akthar Khan, Kumar Devasish, A. Chitra, and Ryan Sujith Electric Vehicles Integration with Renewable Energy Sources and Smart Grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 G. Sree Lakshmi, Rubanenko Olena, G. Divya, and I. Hunko

x

Contents

Allied Technologies FOPD Controller Using Bee Colony Optimized Reduced Order FOFOPDT Model of Three Interacting Tank Process . . . . . . . . . . . . . . 415 U. Sabura Banu, Abdul Wahid Nasir, and I. Mohamed Shiek Mothi Hardware Security for DSP Circuits Using Key-Based Time-Dependent Obfuscation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 R. Uma Rani, D. Jayanthi, Sudharsan Jayabalan, and Arun Vignesh Assessing the Performance of Response Surface Methodology on the Transesterification Process of Moringa Seed Bio-oil . . . . . . . . . . 445 V. Hariram, H. K. Ganesh, Patan Ahmad, Mallela Krishna Kumar, Diviti Somasekhar, S. Seralathan, and T. Micha Premkumar Optimizing the Transesterification Process of Hemp Seed Bio-oil Using Artificial Neural Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 V. Hariram, P. M. Bharadwaj, A. Viswaksen, C. H. Surya, D. Ruthvin Maheej, S. Seralathan, and T. Micha Premkumar Development and Performance Evaluation of Micromagnesium Oxide Filled Silicone Rubber for High Voltage Insulation . . . . . . . . . . . . . . . . 473 Vinayak. V. Rao, Murthy K. Ramakrishana, Pradipkumar Dixit, R. Harikrishna, M. Anand, R. T. Prajwal Kumar, K. M. Raghavendra, and I. S. Sanjay Performance Evaluation of Naturally Aged Silicone Rubber Polymeric Insulating Specimen with Nano TiO2 Filled Epoxy Coating . . . . . . . . . . 485 K. Ramakrishana Murthy, Vinayak. V. Rao, Pradipkumar Dixit, R. Harikrishna, Tarun Anand, Chandrakala, Prajwal Chowdry, and Shubham Kumar Ontology-Based Crime Investigation Process . . . . . . . . . . . . . . . . . . . . . 497 P. Venkata Srimukh and S. Shridevi Variation-Tolerant In-Memory Digital Computations Using SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 D. Gracin, Kichchannagari Omkar, V. Ravi, and T. Vigneswaran Defect Detection and Defect-Tolerant Design of a Multi-port SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 R. Sushmitha, O. R. Nisha, V. Ravi, and T. Vigneswaran

About the Editors

Dr. Pierluigi Siano (M’09–SM’14) received the M.Sc. degree in electronic engineering and the Ph.D. degree in information and electrical engineering from the University of Salerno, Salerno, Italy, in 2001 and 2006, respectively. He is a Professor and Scientific Director of the Smart Grids and Smart Cities Laboratory with the Department of Management & Innovation Systems, University of Salerno. His research activities are centered on demand response, on the integration of distributed energy resources in smart grids and on planning and management of power systems. He has co-authored more than 450 papers including more than 250 international journal papers that received more than 8400 citations with an H-index equal to 46. He received the award as 2019 Highly cited Researcher by ISI Web of Science Group. He has been the Chair of the IES TC on Smart Grids. He is Editor for the Power & Energy Society Section of IEEE Access, IEEE Transactions on Industrial Informatics, IEEE Transactions on Industrial Electronics, Open Journal of the IEEE IES and of IET Renewable Power Generation. Dr. K. Jamuna is an associate professor in the School of Electrical Engineering, Vellore Institute of Technology, Chennai, India. She received her Ph.D from the Indian Institute of Technology Madras in the area of power system, M.Tech. in power systems engineering from College of Engineering Trivandrum, India, and B.E Degree in electrical and electronics engineering from Thiagarajar College of Engineering, Madurai Kamaraj University, Madurai, India. She has published articles in many international journals and conferences, and has also been a reviewer in many international journals. Her research interests include smart grid, power system state estimation, wide area measurement systems and control.

xi

Sustainable Energy Technologies

Z-Source Inverter-Based DVR Using Wind Generator as a Source for Storage Unit Durai Babu and P. Murugesan

Abstract Voltage sag in a electrical distribution system is one of the important tasks and has increased the attention. To compensate the voltage sag, dynamic voltage restorer is used. Due to advanced power electronic technology, compensation capability has increased. DVR using wind generator has taken to prevent the voltage sag and damage for voltage sensitive devices. An inverter is proposed here to maintain the voltage level within the limits and restores the voltage quickly. Shoot through time of the Z-source inverter makes sure the voltage boost when voltage sag occurs. Finally, the developed topology gives the better result and voltage compensation done in a stipulated time and harmonics level of the inverter is also reduced. Keywords Power quality · DVR · Z-source inverter · Wind generator · MATLAB/simulink

1 Introduction Over the last decade and in the twenty-first century, the usage of voltage sensitive devices has been increased and power quality degradation occurs. Such power quality problems can cause severe disturbances and result in power losses. The concept of power quality [1] proposed to provide good quality of power supply, and among recently developed custom power devices, the dynamic voltage restorer is gaining acceptance [2–4]. DVR is preferred over other compensation devices [5] for compensating voltage sag. Real power can be provided by any energy storage device [6] and auxiliary supply energy devices are used in some DVR [1]. An inverter generates the reactive power and Z-source inverter operation and dynamic voltage restorer using wind generator as a storage unit are explained in the following sections.

D. Babu (B) Sathyabama University, OMR, Chennai, India e-mail: [email protected] P. Murugesan S.A. Engineering College, Ponnammallee, Chennai, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_1

3

4

D. Babu and P. Murugesan Line Imped-

Vd Series

Supply

V

V

Transformer

V Z - Source

Load

Inverter with Wind Generator set

Storage Unit

Filter Circuit

Fig. 1 Circuit model of the DVR

Over the last decade and in the twenty-first century, the usage of voltage sensitive devices has been increased and power quality degradation occurs. Such power quality problems can cause severe disturbances and result in power losses. The concept of power quality [1] proposed to provide good quality of power supply, and among recently developed custom power devices, the dynamic voltage restorer is gaining acceptance [2, 3, 4]. DVR is preferred over other compensation devices [5] for compensating voltage sag. Real power can be provided by any energy storage device [6] and auxiliary supply energy devices are used in some DVR [1]. An inverter generates the reactive power and Z-source inverter operation and dynamic voltage restorer using wind generator as a storage unit are explained in the following sections.

2 DVR Operation Figure 1 shows the diagram of dynamic voltage restorer. As shown in Fig. 1 Vs is the supply voltage and Vs1 is the incoming supply voltage, V 2 voltage across load. V dvr is the voltage injected and I is line current. In DVR, there is a wind generator set and battery used a storage unit and the energy from the storage unit is given to the Z-source inverter which converts DC to AC and series transformer is used to which the voltage is injected when voltage sag occurs [2, 3].The inverter output is filtered using the filters to minimize harmonics [7].

3 Z-Source Inverter In traditional method, they used voltage source inverter (VSI) in dynamic voltage restorer to reduce voltage sag but the traditional inverter has no shoot through time and false triggering of switches will lead to damage of inverter. Moreover, only buck

Z-Source Inverter-Based DVR Using Wind Generator …

5

Fig. 2 Circuit model of the DVR

property of the inverter has shown least voltage sag reduction [5, 8–10]. This added boost converter to minimize this drawback [11]. This results in increase in switching devices and more harmonics, and so, they used additional transformer and bridge rectifier later [12]. To overcome all these limitations, they proposed Z-source inverter [8]. The Z-source inverter use L and C network thereby turn on off same leg switches is possible and this we call as shoot through state. We proposed Z-source inverter where wind generator is used as a storage unit and the energy generated by wind generator is used during the voltage sag and shoot through state of the inverter used for compensation.

4 Storage Unit The storage unit capacity becomes very important when we compensate the voltage sag. In this, wind generator is used as storage to overcome this drawback [11]. By using wind generator, DC link voltage will not decrease when compensating the voltage sag. Figure 2 shows the proposed wind generator used as a source for storage unit. Pulse width modulation technique is used here and the shoot through time of the Z-source inverter is calculated and inverter will be operated in shoot through state whenever voltage sag occurs so dynamic voltage restorer can effectively compensate the voltage sag [9] (Fig. 3).

5 Controller Design   1 t ∫ verror (τ )dτ P(t) = K c 1 + Ts 0 where t Ts Vdc Kc

Time at each iteration, Sample time interval. DC voltage. Controller gain.

Calculation of V s and T sh : Step-1: Convert the voltages into stationary frames, i.e., abc to dq0 transformation

6

D. Babu and P. Murugesan

Fig. 3 Switching pattern

     2 2∅ 2∅ va ∗ sin(ωt) + vb ∗ sin ωt − + vc ∗ sin ωt + 3 3 3      2 2∅ 2∅ va ∗ cos(ωt) + vb ∗ cos ωt − + vc ∗ cos ωt + vq = 3 3 3 vd =

v0 = where ∅→ va , vb and vc ω t

180°. Output voltages. Phase angle. Time period

1 (va + vb + vc ) 3

Z-Source Inverter-Based DVR Using Wind Generator …

7

Fig. 4 Circuit model of ZSI with filter

Vs =



Tsh =

Vd − Vq

2

 21  + 2 Vd Vq

k(k + 2)(k − 1) − 1  Ts  (k + 1) (2k)2 − 1

where Vs Tsh k Ts

Switching voltage. Shoot through time. Duty ratio. Switching period.

6 Simulation Results 6.1 Three-Phase Output Voltage at Grid Side The three-phase voltage (ABC) obtained from the nonlinear load is converted into the dq0 references by applying the dq0 transformation technique. Then, the DC quantities are estimated from the dq0 transformation, and it is given to the high pass filtering technique that reduces the harmonic contents. The filtered outputs of the DC components for phase A, phase B, and phase C are represented in Figs. 4, 5, 6, and 7.

6.2 SPWM Pulse Triggered Three-Phase Output Voltage at ZSI After generating the pulses for inverter, the three-phase load current is given to the nonlinear load as shown in Fig. 8. In which, yellow color line indicates phase A, pink color indicates phase B, and blue color indicates phase C. Then, the phase amplitude

8

D. Babu and P. Murugesan

Fig. 5 Voltage waveforms of the ZSI with filter

Fig. 6 Current waveforms of the ZSI with filter

of the output voltages is maintained as constant for each phase by analyzing the depth of voltage sag (Figs. 9, 10 and 11).

Z-Source Inverter-Based DVR Using Wind Generator …

9

Fig. 7 FFT analysis for the ZSI current with filter

Fig. 8 Circuit for line model without DVR system

7 Conclusion Dynamic voltage restorer using wind generator as a storage unit has been implemented and its performance for compensating the voltage sag shows good results. Z-source inverter is used in shoot through state and shoot through time of the Z-source inverter is calculated. This is implemented in this DVR and checked the compensation capability of the device. From the results, it is shown the proposed topology

10

D. Babu and P. Murugesan

Fig. 9 Simulated results of voltage waveform across load without DVR system in the line model

Fig. 10 Circuit for line model with DVR system

Z-Source Inverter-Based DVR Using Wind Generator …

11

Fig. 11 Simulated results of voltage waveform across load with DVR system in the line model

shows the good result by reducing the harmonics generated by the inverter and fast response of the inverter for compensation of voltage sag. Wind generator used as storage unit and DC link voltage has been increased by using this as a storage unit.

References 1. Hingorani NG (2006) Digital simulation of DVR for voltage mitigation by Nor and Mohammed. IEEE Trans Power Electron 2. Samra NA, Neft C, Sundaram A, Malcolm W (1995) The distribution system dynamic voltage restorer and its applications at industrial facilities with sensitive loads. Proc. Power Conversion Intell. Motion Power Quality Long Beach, CA 3. Choi SS, Li BH, Vilathgamuwa DM (2000) Dynamic voltage restoration with minimum energy injection, IEEE Trans. Power Syst 15:51–57 4. Woodley NH, Morgan L, Sundaram A (1999) Experience with aninverter-based dynamic voltage restorer in IEEE Trans. Power Deliv 14:1181–1186 5. Wang B, Venkataramanan G (2004) Evaluation of shunt and series power conditioning strategies for feeding sensitive loads. In: Conference of Record IEEE power electronics conference, vol 21, July 2004 6. Loh PC, Vilathgamuwa DM, Lai YS, Chua GT, Li Y (2006) Voltage sag compensation with z-source inverter based dynamic voltage restorer. In: Industrial Application, vol 5 7. Power Electronic Control in Power System by Aacha 8. Peng FZ (2003) Z-source inverter IEEE Trans. Ind Appl 39:504–510 9. Loh PC, Vilathgamuwa DM, Lai YS, Chua GT, Li Y (2004) Pulse-width modulation of Zsource inverters. In: Conference of record IEEE industry applications conference, 39th IAS Annual Meeting 10. Nor, Mohammed (2006) Digital Simulation of DVR for Voltage mitigation. IEEE Trans Power Electron

12

D. Babu and P. Murugesan

11. Sng EKK, Choi SS, Vilathgamuwa DM (2004) Analysis of series compensation and DC-link voltage controls of a transformer less self charging dynamic voltage restorer in IEEE Trans. Power Deliv 19:1511–1518 12. Jimichi T, Fujita H, Akagi H (2005) Design and experimentation of adynamic voltage restorer capable of significantly reducing an energy storage element. In: Conference of record industry applications conference, 14th IAS Annual Meeting

Predicting Wind Turbine Blade Fault Condition to Enhance Wind Energy Harvest Through Classification via Regression Classifier A. Joshuva, M. Arjun, R. Murugavel, V. A. Shridhar, G. S. Sriram Gangadhar, and S. S. Dhanush Abstract Wind energy has turned into a huge contender of usual fossil fuel energy. The advancement of substantial wind turbines empowers to obtain energy more proficiently as a result of the growing interest for renewables on the planet. With the expanded zest for the usage of wind turbine power plants in remote ranges, basic condition monitoring will be one of the main factors in the proficient foundation of wind turbines in the energy field. The wind turbine is utilized to change over wind energy into electrical energy. To make wind energy more engaged from various resources of energy, related to execution, convenience, dependability, viability, the life of turbines must be enhanced. Fault recognition on cutting edge at an early stage will avoid the issue, as sharp edge destruction can prompt a disastrous result for the whole wind turbine framework. This paper brings a pattern recognition technology into the wind energy field and endeavours to anticipate a different fault condition which happens in wind turbine sharp edge using vibration signals.

A. Joshuva (B) · M. Arjun · G. S. Sriram Gangadhar · S. S. Dhanush Centre for Automation & Robotics (ANRO), Hindustan Institute of Technology and Science, Chennai, Tamil Nadu 603103, India e-mail: [email protected] M. Arjun e-mail: [email protected] G. S. Sriram Gangadhar e-mail: [email protected] S. S. Dhanush e-mail: [email protected] R. Murugavel Information Technology Manager, Global Technology and Business Centre (GTBC), Ford Motor Pvt. Ltd, Chennai, Tamil Nadu 60011, India e-mail: [email protected] V. A. Shridhar Department of Automobile Engineering, Hindustan Institute of Technology and Science, Chennai, Tamil Nadu 603103, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_2

13

14

A. Joshuva et al.

Keywords Condition monitoring · Statistical features · Wind turbine · Wind energy harvest · Classification via regression · Ten-fold cross validation

1 Introduction Wind energy is now turned into the most sought-after resource on the basis of general natural taint [1]. Wind energy is one of the most efficient renewable sources and an alternative to commonly used sources. The wind turbine is used to convert wind energy into electrical energy. The life of turbines must be enhanced [2] in order to make wind energy more engaged from different energy resources, related to execution, convenience, reliability, viability. The airflow makes a rotor pole position switch, allowing the shaft inside the wind turbine shift. The soft, long, and elastic sharp edge of the component has the most extraordinary dynamic strength. It is easy to cause blade vibration [3]. The regular and intense movement would bring destruction to the rim of the rotor and, lastly, to catastrophe. The rim of the knife is one of the critical sections of repeated failures. Due to its gigantic development and working condition, the blade vibration is difficult to find online. In this work, failure discovery on the cutting edge of the wind turbine was performed using vibration data and classification via regression classifier. Many research and study were done on fault examination of cutting edge. To give some case in point, Chen et al. [4] led an analysis on wind turbine pitch issues utilizing from the earlier learning-based ANFIS utilizing SCADA information and got 88.30% accuracy. Mollineaux et al. [5] have done a work on fault recognition strategies on wind turbine sharp edge testing with wired and remote accelerometer sensors utilizing benchmark information and autoregressive moving normal (ARMA) and continuous wavelet transform (CWT) utilized as demonstrating systems. Benim et al. [6] did a work on streamlining of airfoil profiles for little wind turbines with accentuation on stable execution under exceptionally wind conditions utilizing Mesh Adaptive Direct Search (MADS) enhancement approach and got 39% of forecast exactness. Frost et al. [7] have done an examination on integrating structural health management with contingency control for wind turbines utilizing nonlinear high fidelity simulation and accomplished 90% precision in their work. A relative report on wind turbine control coefficient estimation by soft computing approaches was completed by Shamshirband et al. [8]. In this examination, they utilized support vector regression (radial basis function), support vector regression (polynomial), adaptive neuro-fuzzy inference system (ANFIS), NN (neural network) calculations for correlation. The correlation coefficient of calculations was observed to be SVR (RBF)-0.997, SVR (Polynomial)-0.504, ANFIS-0.978, NN0.922. Godwin and Matthews [9] done work on wind turbine pitch faults through SCADA information investigation and RIPPER classifier which yield them 87.05% precision in pitch fault analysis. Various works were done utilizing simulation work and not many in experimental examinations [10]. Machine learning (AI) approach

Predicting Wind Turbine Blade Fault Condition to Enhance Wind …

15

Fig. 1 Methodology

was less considered for wind turbine cutting edge condition investigation and exceptionally less faults were considered. This examination makes an undertaking to find different cutting edge defects using AI approach. Figure 1 demonstrates the technique of the work done.

2 Experimental Studies The trial arrangement, deficiencies, and working technique are explained in detail in Joshuva and Sugumaran [11]. The sampling frequency used in the assessment was 12,000 Hz and each signal has a length of 10,000 data. Piezoelectric accelerometer (DYTRAN 3055B1) was used close by DAQ (NI-USB 4432) for picking up data. For each condition of the front line, 100 example information bundles were taken.

16

A. Joshuva et al.

3 Feature Extraction Using Statistical Analysis For better and other faulty states of a blade, the vibration signal was achieved. If the time domain evaluated signal is taken for a classifier, then the number of tests should be consistent at that point. The number of samples is reached for the revolution of the cutting edge velocity. Therefore, the various speeds correspond. Therefore, the capability of digitized signal data points is too high. Classifiers’ function do typically not handle it successfully. Therefore, formerly the classification system must be derived out for a specific feature [12].

4 Feature Selection Using J48 Algorithm The Classifier J48 in WEKA C4.5 [13–15] is accustomed. This includes different divisions, one base, separate hubs, and specific leaves. A branch is a root to a leaf chain of hubs, each of which has a single quality. Information on the value of the relative quality [16] is given in the event of a trait in a line. The decision tree is a tree-based system of grouping data. The identification of J48 is generally used to build decision trees [17–20]. This highlight function has been linked to the J48 tree classifier. The contribution to the category is the arrangement of facts mentioned above and the decision tree’s production was shown in Fig. 2 [21]. In the centrality request, different features in the nodes of the decision tree work. It should be noticed here that the single elements that relate to the structure are not very significant in the

Fig. 2 J48 tree classification for feature selection

Predicting Wind Turbine Blade Fault Condition to Enhance Wind …

17

decision-making tree. The less distinguished features can be deliberately responsible for by choosing the threshold. This principle is used to pick good characteristics. The analysis identifies the great characteristics with the final objective of organizing the data index and knowledge needed to pick a better fault characteristic [22]. Figure 2 reference, the sum range standard deviation, and kurtosis can be defined as the most dominant features to reflect the blade conditions.

5 Classification via Regression Classifier The category used by regression strategies for the quality of the assignment is called classification via regression [23]. The category is binarised and for each group, there is a single regression model built. Regression measures the relationships among a dependent variable and one or more independent variables in each category. The regression category is as follows [24]. • Assume ignoring the target output y is binary (0 or 1) relatively than a continuous variable. • Then the estimation of linear regression function is supposed to be

f (x; w) = w0 + w1 x1 + · · · + wd xd

(1)

f (x; w) = w0 + w T w1

(2)

• which depends on the data which existing  before the regression. • Assuming y = f (x; w)+ ∈, ∈∼ N 0, σ 2 , then the machine learning objective for the factors w reduces to least squares fitting Jn (w) =

n 1 (yi − f (xi ; w)2 n i=1

(3)

• Then the resulting regression function is found to be   f x; wˆ = w0 + w T wˆ 1

(4)

6 Results and Discussion In good condition, the cutting edge and other faulting conditions of the cutting edge of the wind turbine are noted. In all 600 examples accumulated, 100 of these examples had a clear cutting edge in good condition. Hundred instances from each case are

18

A. Joshuva et al.

Table 1 Confusion matrix of classification via regression Blade conditions

Good

Good

83

Bend 0

Bend

0

Crack

0

Loose

Crack

Loose

1

16

87

8

7

89

14

0

Pitch twist

0

Erosion

0

Pitch twist

Erosion

0

0

0

0

5

4

0

0

4

82

0

0

0

0

0

97

9

2

0

0

3

95

Table 2 Classwise accuracy of classification via regression Class

TP rate

FP rate

Precision

Recall

F-measure

ROC area

Good

0.83

0.028

0.085

0.83

0.843

0.975

Bend

0.87

0.018

0.906

0.87

0.888

0.99

Crack

0.89

0.026

0.873

0.89

0.881

0.989

Loose

0.82

0.04

0.804

0.82

0.812

0.97

Pitch twist

0.97

0.006

0.97

0.97

0.97

0.998

Erosion

0.95

0.016

0.95

0.95

0.936

0.994

accrued for different issues such as blade bend, erosion, blade crack, hub-blade loose connection, pitch angle twist [25]. The numerical properties have been categorized into usability and lead to the category. The corresponding data state is the calculation’s simple performance [26]. The errors were conceived and connected to the good signal to detect whether or not the cutting edge is faulty and to test the quality of the classifier for ten-fold cross validation [27]. The corner-to-corner characteristics of the uncertainty matrix (Table 1) indicate the well-defined instances. The classification accuracy of the classification via regression was 89%. The value of kappa in this category was 0.866 and average absolute error of 0.0764 was observed [28]. The average root error of 0.1742 is less. Table 2 provides comprehensive classwise accuracy of classification via regression. For a superior classifier, the true positive (TP) value should be similar to 1 and the false positive (FP) rate about 0 [29]. The TP rate in the vast majority of groups is below 1 and the FP value is almost 0 can be seen in Table 2 [30].

7 Conclusion The wind turbine is noteworthy in the creation of wind vitality in everyday life. This paper shows a machine learning elucidation of vibration patterns for the valuation of wind turbine cutting edge conditions. The achieved vibration information is modelled utilizing data mining procedure. Classification via regression classifier was utilized

Predicting Wind Turbine Blade Fault Condition to Enhance Wind …

19

to study about and group the various issues of the cutting edge. The model is tried under ten-fold cross validation and appropriately classified instances were observed to be 89%. The error-rate is nearly less and might be considered for the sharp edge deficiency determination. Consequently, the classification via regression classifier can be basically utilized for the condition checking of wind turbine sharp edge to diminish the downtime and to harvest more wind vitality.

References 1. Joshuva A, Sugumaran V (2019a) Crack detection and localization on wind turbine blade using machine learning algorithms: a data mining approach. Struct Durab Health Monit (SDHM). 13(2):181–203 2. Joshuva A, Aslesh AK, Sugumaran V (2019) State of the art of structural health monitoring of wind turbines. Int J Mech Prod Eng Res Develop 9(5):95–112 3. Joshuva A, Sugumaran V (2017a) A comparative study of Bayes classifiers for blade fault diagnosis in wind turbines through vibration signals. Struct Durab Health Monit (SDHM). 12(1):69–90 4. Chen B, Matthews PC, Tavner PJ (2013) Wind turbine pitch faults prognosis using a-priori knowledge-based ANFIS. Expert Syst Appl 40(17):6863–6876 5. Mollineaux M, Balafas K, Branner K, Nielsen P, Tesauro A, Kiremidjian A, Rajagopal R (2014) Damage detection methods on wind turbine blade testing with wired and wireless accelerometer sensors. In: EWSHM-7th European workshop on structural health monitoring 6. Benim AC, Diederich M, Nikbay M (2015) Optimization of airfoil profiles for small wind turbines. In: 8th ICCHMT, Istanbul, 25–28 May 2015 7. Frost SA, Goebel K, Obrecht L (2013) Integrating structural health management with contingency control for wind turbines. IJPHM Special Issue on Wind Turbine PHM ´ 8. Shamshirband S, Petkovi´c D, Saboohi H, Anuar NB, Inayat I, Akib S, Cojbaši´ c Ž, Nikoli´c V, Kiah ML, Gani A (2014) Wind turbine power coefficient estimation by soft computing methodologies: comparative study. Energy Convers Manage 81:520–526 9. Godwin JL, Matthews P (2013) Classification and detection of wind turbine pitch faults through SCADA data analysis. Int J Prognost Health Manage 4(2–16):1–11 10. Joshuva A, Sugumaran V (2016) Fault diagnostic methods for wind turbine: a review. Asian Research Publishing Network (ARPN) J Eng Appl Sci 11(7):4654–4668 11. Joshuva A, Sugumaran V (2017b) A data driven approach for condition monitoring of wind turbine blade using vibration signals through best-first tree algorithm and functional trees algorithm: a comparative study. ISA Trans 31(67):160–172 12. Manju BR, Joshuva A, Sugumaran V (2018) A data mining study for condition monitoring on wind turbine blades using Hoeffding tree algorithm through statistical and histogram. Int J Mech Eng Technol 9(1):1061–1079 13. Joshuva A, Sugumaran V (2018a) A study of various blade fault conditions on a wind turbine using vibration signals through histogram features. J Eng Sci Technol. 13(1):102–121 14. Joshuva A, Sugumaran V (2019b) Selection of a meta classifier-data model for classifying wind turbine blade fault conditions using histogram features and vibration signals: a data-mining study. Prog Ind Ecol Int J 13(3):232–251 15. Joshuva A, Sugumaran V (2019c) Improvement in wind energy production through condition monitoring of wind turbine blades using vibration signatures and ARMA features: a data-driven approach. Prog Ind Ecol Int J 13(3):207–231 16. Joshuva A, Sugumaran V (2017c) Classification of various wind turbine blade faults through vibration signals using hyperpipes and voting feature intervals algorithm. Int J Perform Eng 13:247–258

20

A. Joshuva et al.

17. Joshuva, A., Sugumaran, V. A comparative study for condition monitoring on wind turbine blade using vibration signals through statistical features: a lazy learning approach. Int J Eng Technol 7(4–10):190–196. 18. Joshuva A, Sugumaran V (2018b) A machine learning approach for condition monitoring of wind turbine blade using autoregressive moving average (ARMA) features through vibration signals: a comparative study. Progress in Industrial Ecology, an International Journal. 12(1– 2):14–34 19. Joshuva A, Sugumaran V (2019d) Comparative study on tree classifiers for application to condition monitoring of wind turbine blade through histogram features using vibration signals: a data-mining approach. Struct Durab Health Monit (SDHM) 13(4):399–416 20. Joshuva A, Sugumaran V (2019e) A lazy learning approach for condition monitoring of wind turbine blade using vibration signals and histogram features. Measurement 23:107295 21. Joshuva A, Deenadayalan G, Sivakumar S, Sathishkumar R, Vishnuvardhan R (2019) Implementing rotation forest for wind turbine blade fault diagnosis. Int J Recent Technol Eng 8(2 Special Issue 11):185–192 22. Joshuva A, Vishnuvardhan R, Deenadayalan G, Sathishkumar R, Sivakumar S (2019) Implementation of rule based classifiers for wind turbine blade fault diagnosis using vibration signals. Int J Recent Technol Eng 8(2 Special Issue 11):320–331 23. Loh WY (2011) Classification and regression trees. Wiley Interdiscipl Rev Data Mining Knowled Dis 1(1):14–23 24. Frank E, Wang Y, Inglis S, Holmes G, Witten IH (1998) Using model trees for classification. Mach Learn 32(1):63–76 25. Joshuva A, Deenadayalan G, Sivakumar S, Sathishkumar R, Vishnuvardhan R (2019) Logistic model tree classifier for condition monitoring of wind turbine blades. Int J Recent Technol Eng. 8(2 Special Issue 11):202–209 26. Jung K, Bae DH, Um MJ, Kim S, Jeon S, Park D (2020) Evaluation of nitrate load estimations using neural networks and canonical correlation analysis with K-fold cross-validation. Sustainability. 12(1):400 27. Joshuva A, Sivakumar S, Vishnuvardhan R, Deenadayalan G, Sathishkumar R (2019)Research on hyper pipes and voting feature intervals classifier for condition monitoring of wind turbine blades using vibration signals. Int J Recent Technol Eng 8(2 Special Issue 11):310–319. 28. Braun T, Spiliopoulos S, Veltman C, Hergesell V, Passow A, Tenderich G, Borggrefe M, Koerner MM (2020) Detection of myocardial ischemia due to clinically asymptomatic coronary artery stenosis at rest using supervised artificial intelligence-enabled vectorcardiography–a five-fold cross validation of accuracy. J Electrocardiol 29. Joshuva A, Sivakumar S, Sathishkumar R, Deenadayalan G, Vishnuvardhan R (2019) Fault diagnosis of wind turbine blades using histogram features through nested dichotomy classifiers. Int J Recent Technol Eng 8(2 Special Issue 11):193–201 30. Xiong Z, Cui Y, Liu Z, Zhao Y, Hu M, Hu J (2020) Evaluating explorative prediction power of machine learning algorithms for materials discovery using k-fold forward cross-validation. Comput Mater Sci 1(171):109203

Satin Bower Bird Algorithm for Controller Parameter Optimization in an Autonomous AC Microgrid Rekha P. Nair and P. Kanakasabapathy

Abstract Satin bower bird optimizer (SBO) is a very recent meta-heuristic algorithm used in optimization problems. In this work, an investigation is done on the efficacy of SBO in tuning a PI controller parameter for improving voltage stability in an autonomous AC microgrid. Intelligent microgrids are in fact the building blocks of a smart grid. For a microgrid operating in the autonomous mode, maintaining frequency and voltage stability under different disturbance conditions are major causes of concern. Droop control is the most accepted control for ensuring proper current sharing among parallel converters. The AC bus voltage is affected by the change in reactive power demand, and maintaining the bus voltage stability by minimizing the voltage deviation error is studied in this work. Here, a PI controller is employed whose gain is optimally tuned by SBO. The effectiveness is further tested by comparing the performance with controller gain set to values above and below the optimally tuned values. Keywords Ac microgrid · Optimization · Satin bower bird · Droop control · Voltage stability · Steady state error

1 Introduction Modern electric power system has developed to the concept of microgrid which is the combination of energy storage, loads, inverters, distribution generators, and monitor devices. Mahmud and Town [1] is a review paper which considers many simulation tools that have been employed for modeling power systems for specific tasks. The different modes of operation were explained briefly in Pogaku et al. [2]. In grid-connected mode, the inverters are needed to generate the active as well as reactive power and currents, which ensure the high quality of output power levels. In autonomous mode, the frequency and voltage value is predefined with the help of R. P. Nair (B) · P. Kanakasabapathy Department of Electrical and Electronics Engineering, Amrita Vishwa Vidyapeetham, Amritapuri, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_3

21

22

R. P. Nair and P. Kanakasabapathy

control strategy. The power sharing strategy can vary from centralized to decentralized manner in the islanded microgrid. Nowalid [3] utilizes the small signal model for analyzing a dynamic microgrid The principle control techniques like micro-generator controller, central controller, and load controllers are defined in [4–7]. Syrri et al. [8] describes the utilization of ant lion optimization for various engineering applications. Dobakhshari et al. [9], and Anoop et al.[10] deals with various control aspects of an islanded microgrid and its optimal scheduling of DG’s. Bharath et al. [11] reviews control strategies for a DC microgrid. There are a wide variety of optimization problems in the operation of a microgrid. For example, the optimal setting of droop coefficients for efficient current sharing, minimization of cost, the optimal usage of renewable sources, power quality optimization, minimization of voltage fluctuation, and voltage deviation minimization are certain cases of optimization problems in a microgrid. Kai et al. [12, 13] describes the optimization of controller parameters in a microgrid using meta-heuristic algorithms. The applications of SBO in engineering are discussed in [14] and [15]. Hassan and Abido [16] depicts the effectiveness of particle swarm optimization for microgrid application both the grid connected and autonomous mode. In Carlos and Guerrero [17], a clear insight is given into the various optimization problems encountered during the planning process of a microgrid. A novel optimization algorithm is applied on a typical low-voltage microgrid operating under various market policies in Tsikalakis and Hatziargyriou [18]. Different intelligent optimization algorithms namely fuzzy logic, particle swarm optimization, and bacterial search algorithm developed to tune the control parameters are surveyed in [19]. Kumar and Kalavathi [20] presents a novel strategy based on satin bower bird optimization to obtain the optimal sizing and power management of hybrid photovoltaic/wind/battery power system.

2 Proposed System The block diagram for the proposed system is shown in Fig. 1. The system consists of distributed generators that can be fed to inverter, which convert DC to AC, and then, it is passed to the LC filter to remove the harmonics presented in the output of the inverter. The power calculating block calculates the active and reactive power. Then, the voltage and frequency are calculated according to the droop control strategy with active and reactive power, and it is fed to the PI controller. In this work, the PI controller is tuned with the SBO to improve the voltage stability. The control of inverters in AC microgrids describes the droop control technique for automatic power sharing among the microsources. The grid interface DC/AC power electronic inverters are paralleled together to form one AC bus which is connected to the grid via a static transfer switch (STS). Local loads can be connected to the microgrid side of the STS. DC/AC inverters are change critical devices of any power electronicbased microgrid consisting of high switching frequency solid-state devices and a

Satin Bower Bird Algorithm for Controller Parameter …

23

Fig. 1 Block diagram of the proposed system

low pass filter, when a voltage source inverter (VSI) is controlled in order to feed the load with prior calculated values for voltage and frequency, according to a specific control strategy such as the power frequency droop control, and depending on the load, the VSI active and reactive power output is defined. The control principle of the VSI in this case emulates the behavior of a synchronous machine. Thus, it is possible to control voltage and frequency on the AC system by means of inverter droop control. The equation governing this droop characteristic is p − po = −m p ( f − f o )

(1)

V − Vo = −n q (Q − Q o )

(2)

whereP and Po refers to the actual and reference values of active power, f and f o are the actual and nominal frequency, mp is the active power droop coefficient given by Eq. (3), and nq is the reactive power droop coefficient given by Eq. (4) (Figs. 2 and 3). mp = nq =

po f max − f min

(3)

Vo Q max − Q min

(4)

24

R. P. Nair and P. Kanakasabapathy

Fig. 2 Active power verses frequency droop

Fig. 3 Reactive power versus voltage droop

Q

3 Satin Bower Bird Optimization Technique Satin bower bird optimization technique (SBO) is a new bio-inspired optimization technique which can be employed for various engineering applications. It is very effective in optimizing and power management. The satin bower bird optimization is bio-inspired by the attitude of the male-mesmerizing-the-female for mating. The male bower bird constructs a specialized bower to attract the female. Male birds make dedicated twig buildings, called bowers, to attract the female. Bowers are adorned with blossoms, quills, berries, and so on. These beatifications are imperative in feminine optimal and male coupling realization. Males contest by robbing adornments from different males and will abolish the neighbor’s bowers. Male wooing activities contain demonstration of beatifications and dancing shows conveyed through noisy communications. An additional signal of resilient sensual opposition is that not entire mature birds are effective at building, continuing, and also protecting bowers. Consequently, there is significant unpredictability in male copulating achievement. Various stage processes involved in the algorithm are shown in Fig. 4.

Satin Bower Bird Algorithm for Controller Parameter …

25

Fig. 4 Stage processes of SBO

3.1 Bower Initialization Random generation of the initial population is the first step. Initial population consists of a set of positions for bowers where each position is defined as a n-dimensional vector of parameters to be optimized. The parameters of each bower are identical with the variables of the optimization problem. As a first step, in the initialization of the bower’s population, gain of the controller is set as a variable with lower and upper bound value as say 10 and 50, respectively. The number of times to repeat the process of detecting the bower among the specified population is also given as input through max iterations. The value of the gain of the controller will lie between 10 and 50 based on its boundary values.

3.2 Calculating Bower Probabilities The bower can be selected as the best bower based on optimizing the cost function with a minimum value. This probability of selecting the best bower among the population of bower is determined by the following equation. fitnessi Probability, Pi =  N i=1 fitnessi

(5)

26

R. P. Nair and P. Kanakasabapathy

where N is the bower number and fitnessi is the fitness of the i-th position. Fitness of each position is calculated using the following equations. fitnessi =

1 1+ f

(6)

where f i is the fitness function which is the magnitude of the voltage deviation. Error in this problem is greater than or equal to 0. Voltage deviation error = V(rms ref) − V(rms actual)

(7)

3.3 Population Updating The positions of the bowers are updated for each iteration by the three parameters. Based on these parameters the bower position is calculated and it is used for calculating the objective function. The three factors are 1. Chances for getting elite bower. 2. Best bower position. 3. Updating the position of bower by roulette wheel selection method. The probability of selecting a bower as a best bower is defined as the chances for getting elite bower. It is calculated as follows: λ=

α Pi

Pi new == Pi old + λ{(Pi + Pelite )/2 − Pi old }

(8) (9)

where α is the greatest step size and Pi is the probability obtained by Eq. (4) The updated position of the bower is given by Eq (8).

3.4 Population Reduction The bower is reduced by the normal distribution process which is based on the boundary of the gain and as well as the elitism. It is helpful for calculating the best bower position or the gain of the controller.

Satin Bower Bird Algorithm for Controller Parameter …

27

3.5 Global Best Bower Detection From the elitism process, the local best bower is determined for each iteration among the total bowers. This local best bower position will be sorted in the ascending order, and the topmost bower is selected as the global best bower. The convergence curve of the process can be obtained when the maximum iteration is reached. The steps of the SBO algorithm are as follows: (a) (b) (c) (d) (e) (f) (g) (h)

Generate a random population of bowers. Evaluate the fitness value f (i) of each bower from the population. Find the best bower and assume it as elite. Calculate the probability of bowers using Eqs. (5) and (6). Update the position of bower using Eqs. (8) and (9). Calculate the fitness of all bowers. Update elite if a bower becomes fitter than the elite. Return the best bower.

4 Implementation of the Proposed System The proposed ac microgrid with SBO optimized PI controller is implemented using MAT LAB Simulink. The SBO algorithm is implemented by MAT LAB coding. The proposed system is simulated for reactive power ranging from 1000 VAR to 1800 VAR. The gain of the PI controller is tuned by the SBO algorithm. The code for SBO algorithm is run and the best value returned is set as the controller gain. The results obtained from simulation are discussed in the next section (Figs. 5, 6 and 7).

Fig. 5 MATLAB simulink model of the proposed system

28

R. P. Nair and P. Kanakasabapathy

Fig. 6 Steady state error of the three phase voltages for the optimized PI controller gain

Fig. 7 Steady state error of the three phase voltages for PI controller gain less than optimal value

5 Simulation Results The optimized gain value for reactive power 1500 VAR was obtained as 26.4718. With this value, the output voltage of the inverter is closely following the reference with a peak value of 340 V. The error plots with different values of PI controller gain are compared with the SBO optimized value and are analyzed in detail. Fig 8 shows the tracking of voltage deviation in the proposed system with SBO optimized PI controller with the reference voltage. The tracking efficiency is compared with two different gain values. At steady state, the error w.r.t the reference voltage of R, Y, B phases are, respectively, found as 1.9, 0.9, and 0.621 V. The optimized controller gain obtained by running SBO is tabulated in Table 1. The validation of the results in terms of performance indices are tabulated in Table 2.

Satin Bower Bird Algorithm for Controller Parameter …

29

Fig. 8 Steady-state error of the three phase voltages for PI controller gain greater than optimal value

Table 1 Controller gain optimized by SBO for different reactive power values Reavtive power values (VAR)

Optimized PI controller gain

1000

26.4718

1400

26.6512

1450

26.6512

1500

26.5981

1800

26.5258

Table 2 Error indices for different controller gains Controller parameter for reactive power (1000 VAR)

Error indices IAE

ISE

Optimal value = 26.4718

0.86

Less than optimal = 20

2.92

8.34

Greater than optimal = 30

5.29

16.76

2.93

6 Conclusion In this work, the voltage profile of the microgrid is tracked and maintained at the desired level by a novel control algorithm. It is a fusion of PI controller-based droop strategy which is optimized by SBO technique. The results of the work prove that the tracking efficiency is considerably increased, and at the same time, the convergence time is also small. The results indicate that the algorithm is able to perform adaptive tuning subject to variations in reactive power. Even though SBO technique has been tried and proven for its effectiveness in various problems including optimal load dispatching, the idea implemented in this paper is a new one.

30

R. P. Nair and P. Kanakasabapathy

References 1. Mahmud K, Town, GE (2016) A review of computer tools for modeling electric vehicle energy requirements and their impact on power distribution networks. Appl Energy 172:337–359 2. Pogaku N, Prodanovic M, Green TC (2007) Modeling analysis and testing of autonomous operation of an inverter-based microgrid. IEEE Trans Power Electron 22(2):613–625 3. Nowalid W (2009) Small signal dynamic study of a microgrid 4. Vadana P, Kottayil SK (2016) Dynamic energy management on a hydro-powered smart microgrid. In: Proceedings of the international conference on soft computing systems. Springer, New Delhi, pp 627–635 5. Barklund E, Pogaku N, Prodanovic M, Hernandez-Aramburo C, Green TC (2008) Energy management in autonomous microgrid using stability-constrained droop control of inverters. IEEE Trans Power Electron 23(5):2346–2352 6. Nair RP, Kanakasabapathy P (2017) Control of a DC microgrid under dynamic load condition. In: 2017 international conference on technological advancements in power and energy (TAP Energy). IEEE, pp 1–6 7. Costa PM, Matos MA (2009) Assessing the contribution of microgrids to the reliability of distribution networks. Electric Power Syst Res 79(2):382–389 8. Syrri ALA, Cesena EM, Mancarella P (2015) Contribution of Microgrids to distribution network reliability. In: 2015 IEEE Eindhoven PowerTech. IEEE, pp 1–6 9. Dobakhshari AS, Azizi S, Ranjbar AM (2011) Control of microgrids: aspects and prospects. In: 2011 international conference on networking, sensing and control. IEEE, pp 38-43 10. Anoop S, Ilango K, Nandagopal JL, Nair MG (2017) Implementation of a load side management algorithm for an islanded microgrid powered by renewable energy sources. In: 2017 international conference on technological advancements in power and energy (TAP Energy). IEEE, pp 1–6 11. Bharath KR, Student AD, Kanakasabapathy P (2017) A simulation study on modified droop control for improved voltage regulation in DC microgrid. In: 2017 International conference on intelligent computing, instrumentation and control technologies (ICICICT). IEEE, pp 314–319 12. Kai Y, Ai Q, Wang S, Ni J, Lv T (2015) Analysis and optimization of droop controller for microgrid system based on small-signal dynamic model. IEEE Trans Smart Grid 7(2):695–705 13. Mirjalili S, Jangir P, Saremi S (2017) Multi-objective ant lion optimizer: a multi-objective optimization algorithm for solving engineering problems. Appl Intell 46(1):79–95 14. Moosavi SHS, Bardsiri VK (2017) Satin bowerbird optimizer: A new optimization algorithm to optimize ANFIS for software development effort estimation. Eng Appl Artif Intell 60:1–15 15. Mostafa MA, Abdou AF, Abd El-Gawad AF, El-Kholy EE (2018) SBO-based selective harmonic elimination for nine levels asymmetrical cascaded H-bridge multilevel inverter. Australian J Electri Electron Eng 15(3):131–143 16. Hassan MA, Abido MA (2010) Optimal design of microgrids in autonomous and gridconnected modes using particle swarm optimization. IEEE Trans Power Electron 26(3):755– 769 17. Carlos G, Guerrero JM (2015) Computational optimization techniques applied to microgrids planning: a review. Renew Sustain Energy Rev 48:413–424 18. Tsikalakis AG, Hatziargyriou ND (2011) Centralized control for optimizing microgrids operation. In: 2011 IEEE power and energy society general meeting. IEEE, pp 1-8 19. Mahmoud MS, Alyazidi NM, Abouheaf MI (2017) Adaptive intelligent techniques for microgrid control systems: a survey. Int J Electr Power Energy Syst 90:292–305 20. Kumar KR, Kalavathi MS Optimal sizing of grid connected hybrid PV/wind/battery power system using satin bowerbird optimization.

Modeling and Simulation of a DFIG-Based Wind Energy System Madhuvanthani Rajendran and L. Ashok Kumar

Abstract Developing nations like India are rapidly shifting toward renewable energy to accelerate its goals toward embracing renewables. Wind power plays a significant role in the changeover to an environmentally friendly future, encouraging extensive research in this sector. India is a key contestant in the global wind energy market and ranks fifth in the world in terms of installed wind power capacity. Currently, India is handling the world’s largest renewable energy program with the target of 175 GW renewable energy by 2022. In this paper, a Doubly Fed Inductionbased wind energy system is simulated using the PLECS software with the analytical calculations being done in MATLAB. The DFIG is preferred in this system, due its capability of controlling as well as generating active and reactive power independently, with the ability to deliver this reactive power to the stator through the grid-side converter when required. The DFIG model is simulated along with the wind turbine model as well as the rotor-side controller and the grid-side converter using the average model of the converters. The DFIG wind generator is simulated in the dq reference frame and the performance of the system is analyzed at various wind velocities. The developed DFIG-based wind energy system is connected to a grid voltage of 575 V, 60 Hz and this voltage is maintained in spite of varying wind velocities which establishes the stability of the system. The analytical calculations done for the various wind velocities are compared with the simulated results and are found to be equivalent. Keywords Doubly-fed induction generator (DFIG) · Wind energy conversion system (WECS) · Distributed generation (DG) · Rotor-side converter · Grid-side converter · Wind turbine control

M. Rajendran (B) Sri Shakthi Institute of Engineering and Technology, Coimbatore, Tamil Nadu 641062, India e-mail: [email protected] L. A. Kumar PSG College of Technology, Coimbatore, Tamil Nadu, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_4

31

32

M. Rajendran and L. A. Kumar

1 Introduction: Wind power has increased exponentially since the dawn of the twenty-first century. As of June 2019, wind power capacity worldwide has reached 597 GW, with 50.1 GW added in the year 2018 [1]. Wind energy can be utilized to generate electricity with the help of wind turbines which convert the kinetic energy present in the wind to mechanical power. The increasing popularity of using wind as an alternate source of energy is due to the environmentally friendly nature and non-polluting properties of wind energy [2]. This paper focuses on the modeling of a Doubly Fed Induction Generator (DFIG)-based wind energy system. The paper is divided into three parts. The first part focuses on the mathematical modeling of the induction machine; the turbine model along with the rotor-side control design is included in the second part and the grid-side control design is included in the third part. Each part is modelled separately using PLECS and the simulations are done for various conditions. Multi-MW wind turbines currently prefer the Doubly fed Induction Generator system [3]. In order to achieve the most favorable aerodynamic efficiency, the aerodynamic system must be efficient in operating over a wide range of wind speeds by tracking the optimum tip-speed ratio [4]. Therefore, the DFIG system operates in both super-synchronous and sub-synchronous modes with a rotor speed range around the synchronous speed since the generator’s rotor must be able to operate at a variable rotational speed. The rotor winding is connected via slip-rings to a three-phase converter while the stator circuit is directly connected to the grid. The DFIG offers adequate performance for variable-speed systems with small speed range requirements such as ± 30% of synchronous speed which is sufficient to exploit the typical wind resources [5]. Doubly Fed Induction Generators have windings on both the rotating and stationary parts, where both the windings transfer significant amount of active power between the shaft and the electrical system [6]. The three-phase rotor winding is fed from the grid through a converter and the stator winding is connected directly to the three-phase grid [7]. An AC–DC–AC converter is incorporated in the induction generator rotor circuit as shown in Fig. 1. The advantage of this configuration is that the power electronic converter needs to be rated only to handle the rotor power which is typically about 30% of the nominal generator power and not the entire power [8]. Therefore, this partially rated power electronic converter leads to lower losses as well as the reduction in the cost of the entire system. The three-phase parameters of the machine such as the voltage, current and flux linkages are transformed into DC values using the park transformation also known as the dq transformation. In order to simplify the analysis of three-phase circuits, this transformation rotates the reference frame of three-phase systems [9]. The application of the dq transform reduces the three AC quantities to two DC quantities for balanced three-phase circuits. The actual three-phase AC quantities can be recovered by performing the inverse transform before which simplified calculations are carried out on these DC quantities. PLECS has a separate block for this transformation. For this simulation, a gain of sqrt(3/2) has been added from abc to dq transformation

Modeling and Simulation of a DFIG-Based Wind Energy System

33

Fig. 1 DFIG model with power converters

after the PLECS block and a gain of sqrt(2/3) for the dq to ABC transformation (grid voltage-oriented control). This paper deals with the modeling, steady-state and dynamic analysis, controller design and simulation of a DFIG-based wind generator.

2 Doubly Fed Induction Machine Modeling In this section, the Doubly Fed Induction Generator model is implemented in PLECS without the wind turbine model [10]. The model is implemented as a squirrel cage machine and this means that the rotor bars are short circuited, and the voltage applied to the rotor windings is zero. Hence, excitation is given only to the stator windings. The stator windings are connected to the grid and a mechanical torque corresponding to a value of 6000 Nm (for motoring action) and −6000 Nm (for generating action) is given to simulate the model. The PLECS schematic corresponding to the DFIG model is given below (Fig. 2). The following system parameters have been considered for the modeling of the DFIG machine. where, Rs is the stator resistance, Rr is the stator referred rotor resistance, L ls and L lr are the stator and stator referred rotor leakage inductances, respectively, L m is the mutual inductance, ω is the reference frame speed, ωr is the rotor electrical speed, Pf is the number of poles of the machine, and C dc is the capacitor of the DC link. The rated power for the machine is S rate = 1.67 MVA, Max. Active power Pmax = 1.5 MW, DC link voltage is V dc = 1150 V, and the grid voltage is 60 Hz, 575 V (line to line RMS value). It has been assumed that the turns ratio of the DFIG is 1:1,

34

M. Rajendran and L. A. Kumar

Fig. 2 PLECS schematic corresponding to the DFIG model

the inertia of the generator is J = 100 kg.m2, and the damping coefficient (B) is 1e–3 Nm/(rad/s). The Doubly Fed Induction Generator works on the principle that two interacting magnetic fields produce a torque. When the stator windings are excited through the grid, current flows in the three windings which are sinusoidally phase shifted by 120° [11]. These currents produce flux in the air gap and due to transformer action currents are being produced in the rotor windings. These currents produce a magnetic field that in turn interacts with the magnetic field produced by the stator currents. This interaction produces electromagnetic torque which is responsible for the rotation of the rotor. The flux linkage equations of the DFIG model are given below [12]. Here, λs . λs denotes the stator flux and λr . λr represents the rotor flux. The subscripts d and q denote the d-axis and the q-axis component, respectively [13]. d [λsd ] = vsd − Rs ∗ i sd + ωd ∗ λsq dt

(1)

d  λsq = vsq − Rs ∗ i sq − ωd ∗ λsd dt

(2)

d [λrd ] = vrd − Rr ∗ i rd + ωd A ∗ λrq dt

(3)

d  λrq = vrq − Rr ∗ i rq − ωd A ∗ λrd dt

(4)

With the values of the stator and rotor flux linkage, it is possible to calculate the stator and rotor currents. The equations of the stator and rotor currents are derived from the equivalent circuit of the DFIG machine in the dq-frame of reference. The values of the required currents are shown in the following matrix.

Modeling and Simulation of a DFIG-Based Wind Energy System

⎤ ⎡ Ls i sd ⎢ i sq ⎥ ⎢ 0 ⎢ ⎥=⎢ ⎣ i rd ⎦ ⎣ L m i rq 0 ⎡

0 Ls 0 Lm

Lm 0 Lr 0

35

⎤−1 ⎡ ⎤ λsd 0 ⎢ ⎥ Lm ⎥ ⎥ ⎢ λsq ⎥ ⎦ ⎣ 0 λrd ⎦ Lr λrq

The electromagnetic torque (T em ) generated by the DFIG is calculated using the currents obtained from the above matrix. T em is obtained from the stator and rotor currents using the following equation. Tem =

P

L m i sq i rd − i sd i rq 2

(5)

In the above equation, P represents the number of poles in the machine and L m represents the mutual inductance of the model.

2.1 Simulation Results for DFIG Machine Modeling: A mechanical torque of 6000 Nm corresponding to motoring action is given to simulate the model and the waveforms for various parameters are extracted here. The torque is stepped down to −6000 Nm corresponding to generating action and the waveforms for the same parameters discussed here along with the inference. From Fig. 3a it can be seen that the electromagnetic torque is 6000.12 Nm which closely matches the analytical value of 6000.4 Nm. Since the DFIG generator without the wind turbine has been simulated in PLECS, the rotor voltages have been set to zero (corresponding to a squirrel cage type operation) with the stator voltages set to rated voltage. Excitation is given only to the stator windings by connecting it to the grid and a mechanical torque corresponding to a value of 6000 Nm (for motoring action) and −6000 Nm (for generating action) is given to simulate the model. It can also be seen from Fig. 3b that the value of rotor speed is 124.574 rad/s which closely matches the analytical value of 124.5737 rad/s. A step input has been applied at 1 s with an initial input of 6000N-m and a final input of —6000 N-m. The following results have been obtained. From Fig. 4a, it can be observed that the step change in torque takes place at 1 s. After an initial disturbance, the torque waveform reaches steady state at around 1.5 s to a value of −6000.3 Nm. This value closely matches the analytical value of − 6000.3 Nm as shown in Table 1. At 1 s, there is a change in rotor speed and after an initial disturbance, it comes to a steady-state value of 126.706 rad/s at around 1.5 s which can be observed from Fig. 4b. This value closely matches the analytical value of 126.7061 rad/s. The rotor and stator currents also attain steady-state values after an initial disturbance. The analytical calculations for the above two cases were done in MATLAB and the results are compared with the simulation results in Table 1 given below: Thus, the analytical and simulation results are found to match with each other.

36

M. Rajendran and L. A. Kumar

Fig. 3 Waveforms for a mechanical torque of 6000 Nm a Electromagnetic torque b Rotor speed c Stator currents d Rotor currents

The following speed—torque curve given in Fig. 5 shows the analytical result of speed pertaining to 6000 and −6000 N-m of torque. The analytical speed values for the two different torques can once again be verified from the following graph.

3 Modeling of DFIG Wind Generator Along with the Wind Turbine Model and Rotor-Side Converter In this section, the DFIG wind generator is modelled including the wind turbine model. This stage comprises of a two-level control. The purpose of turbine control is to extract maximum power when available wind power is below the rated value. One of the methods is to provide power commands as a function of rotor speed. The outer speed controller provides the reference command to the inner current controller [14]. The current controller then translates the current reference commands into control voltage signals that are used in the average model of the rotor-side converter. A wind turbine model is modelled using the following equations.

Modeling and Simulation of a DFIG-Based Wind Energy System

37

Fig. 4 Waveforms for a step change in mechanical torque a Electromagnetic torque b Rotor speed c Stator currents d Rotor currents

Table 1 System parameters Rs ()

Rr ()

L ls (mH)

L lr (mH)

L m (mH)

Pf

C dc (F)

0.0046

0.0032

0.0947

0.0842

1.526

6

0.01

Pmech =

1 ρ Ar vw3 C P (λ, θ ) 2

where, 1 ρ Ar = 2311 2 Vw = Wind speed

c c2 − 5 − c3 θ − c4 e λe C P (λ, θ ) = c1 λ1 (c1 = 0.5167, c2 = 116, c3 = 0.4, c4 = 5, c5 = 21)

(6)

38

M. Rajendran and L. A. Kumar

Fig. 5 Speed—torque characteristics of DFIG for Vr = 0 p.u

1 1 0.035 = − 3 λi λ + 0.08θ θ +1

(7)

where, λ   tip speed ratio. ωmech with K b = 0.539. λ = K b vw θ pitch angle (assumed zero).

3.1 Wind Turbine Model The wind turbine model maintains constant tip speed ratio by varying ωmech . th variation in the wind velocity. The PLECS schematic of the wind turbine model is shown in Fig. 6. The resulting torque from the wind turbine model is given as the load torque to the DFIG model.

3.2 Rotor-Side Converter Under a given operating condition, the main control objective of the rotor-side converter control is to obtain the maximum possible power from the wind. The maximum power tracking characteristics ensure that the rotor always operates at the optimal speed which in turn is controlled by the electromagnetic torque of the

Modeling and Simulation of a DFIG-Based Wind Energy System

39

Fig. 6 PLECS schematic of the wind turbine model

DFIG. The rotor d-axis current controller makes sure that the torque obtained from the speed control loop is maintained at the reference value. The control of rotor injected currents ensures that both the reactive and active powers at the stator can be controlled independently [15]. In this section, the ideal, average model for the converter is used and the DC link is modelled as an ideal voltage source. The PLECS schematic of the rotor-side converter is shown below in Fig. 7. The controllers in this stage were designed using the k-factor method and the MATLAB code for the design is included in the appendix. The plant transfer function of the speed controller is = s Jeq1+B . G p (s) = Tωemr (s) (s) e PLECS schematic of the speed control loop is given below in Fig. 8. The speed reference for this controller is generated from the MPPT curve expression for the wind turbine. The desired operating rotor speed is generated from the speed loop and is in turn used to generate the required electromagnetic torque reference. The obtained torque reference is translated to the current reference using the

Fig. 7 PLECS schematic of the rotor-side converter (average model)

40

M. Rajendran and L. A. Kumar

Fig. 8 PLECS schematic for the speed controller

following equations Ls 1 ∗ 2 ∗ = − ωs T i rd p L m vsd em

(8)

Ls 1 ∗ vsd Qs − L m vsd ωs L m

(9)

∗ i rq =

The d-axis component of the rotor current controls the active power generated and the q-axis component of the rotor current controls the reactive power transferred. The rotor-side current controller is designed using the following equations, vrd = vrq =

di rd Rr i rd + σ L r dt di rq Rr i rq + σ L r dt

− ωd A

Lm λsq − wd A σ L r i rq Ls

(10)

− ωd A

Lm λsd − wd A σ L r i rd Ls

(11)

The plant transfer function of the current controller is G p (s) =

1 i rd (s) = vrd (s) Rr + sσ L r

(12)

G p (s) =

i rq (s) 1 = vrq (s) Rr + sσ L r

(13)

The PLECS schematic of the current controller is shown in Fig. 9. The design of the current controller is done using the k-factor method in MATLAB and a summary of the control design parameters is given below in Table 2. The analytical calculations for two conditions of wind velocity are done in MATLAB and the codes are attached in the appendix. A summary of the results obtained through simulation and analytical calculations is given below in Table 3. It has been observed that the analytical results and the simulation results closely match with each other.

Modeling and Simulation of a DFIG-Based Wind Energy System

41

Fig. 9 PLECS schematic of the current controller

Table 2 Comparison of analytical and simulation results −6000 Nm

6000 Nm T em (Nm) wmech (rad/s)

Analytical

Simulation

6000.4

6000.12

124.5737

Analytical

Simulation

−5999.8

−6000.3

124.574

126.7061

126.706

I sd (A)

1336.3

1336.23

−1286.8

−1286.71

I sq (A)

−1153.5

−1153.46

−1182.8

−1182.73

I rd (A)

−1410

−1409.93

1376.1

1376.62

I rq (A)

236.2525

236.229

246.420

246.528

Table 3 Design parameters for the speed and current controllers S. No.

Controller

ωc (Hz)

Kc

ωz (rad/s)

ωp (rad/s)

1

Current controller

2000

7357.2

3377

46,761

2

Speed controller

1

1057.8

1.6836

23.4491

3.3 Simulation Results of the DFIG Machine with the Wind Turbine and the RSC The model is simulated for two different wind velocities of 12 and 8 m/s and the waveforms for various parameters are extracted here. The waveforms for a ramp in wind velocity from 6 to 12 m/s between 7 and 13 s are also discussed here along with the inference.

42

M. Rajendran and L. A. Kumar

From Fig. 10a, it can be observed that the rotor speed for a wind velocity of 12 m/s is 150.94 rad/s and from Fig. 11a, it can be seen that the rotor speed for a wind velocity if 8 m/s is 106.92 rad/s. The rotor speed is directly proportional to the wind velocity and thus a decrease in the wind velocity leads to a decrease in the rotor speed. The values of the rotor speed for the two different wind velocities match up to the analytical values of 150.8 and 106.92 rad/s, respectively, are as seen in Table 3. The same inference can be made for the grid power as well since the power is also directly proportional to the wind velocity. From Figs. 10b and 11b, it can be observed that the grid power for the wind velocities of 12 m/s and 8 m/s are 1.52 MW and 0.4765 MW, respectively. These values closely match up to the analytical values of 1.4839 MW and 0.52622 MW, respectively. The ramp in the wind velocity from 6 to 12 m/s between 7 and 13 s can be clearly observed from Fig. 12. As already mentioned, since the rotor speed and grid power are directly proportional to the wind velocity, a ramp increase can also be observed in both the quantities as seen in Fig. 12a and b.

Fig. 10 Waveforms for a wind velocity of 12 m/s a Rotor speed b Grid power c Stator currents d Rotor currents e stator voltage f Rotor voltage

Modeling and Simulation of a DFIG-Based Wind Energy System

43

Fig. 11 Waveforms for a wind velocity of 8 m/s a Rotor speed b Grid power c Stator currents d Rotor currents e Stator voltage f Rotor voltage

4 Modeling of the Grid-Side Converter This part of the project deals with the modeling of the grid-side converter (GSC), the current controller for the grid-side converter, and the DC-link voltage controller. This model is then cascaded with the model of the rotor-side converter with the dc-link voltage source replaced by a capacitor of 0.01 F. The converters are simulated using the average model for each of them. The purpose of designing the grid-side converter is to control the transfer of reactive power between the rotor side and the grid and to maintain the voltage of the DC-link capacitor to 1150 V. The grid-side and the rotor-side converters are connected with a DC-link capacitor. The voltage across this capacitor is given as an input to the DC-link controller, from which the grid reference currents are obtained. These currents are fed into the current controller which gives the control voltages for the grid converter. In this section, the detailed model of the current controller and DC-link voltage controller is explained. The system is simulated in PLECS and the performance is studied at wind velocities of 12 and 9 m/s. The results such as the plots of rotor speed, rotor, and grid-side currents are shown.

44

M. Rajendran and L. A. Kumar

Fig. 12 Waveforms for ramp increase in wind velocity a Rotor speed b Grid power c Stator currents d Rotor currents e Stator voltage f Rotor voltage

4.1 DC-Link Voltage Controller The square of the voltage measured from the capacitor link is given as an input to the DC-link controller and the square of the required voltage (1150 V) is subtracted from this term. The difference is then fed to a Type–II controller to obtain the power reference. This is divided by the grid voltage to get the d-axis current reference. The q-axis current reference is dependent on the reactive power requirement. The plant transfer function for the DC-link voltage controller is given by: 2 Vdc2 (s) = Pg (s) sCdc The schematic of the DC-link controller is shown in Fig. 13.

(14)

Modeling and Simulation of a DFIG-Based Wind Energy System

45

Fig. 13 DC-link voltage controller for grid-side converter (GSC)

4.2 Grid-Side Current Controller The control voltages for the average model of the grid-side converter are obtained from this controller. The d-axis and q-axis reference currents are given as inputs to the current controller. The q-axis current is dependent on the reactive power: Q g = eq i d − ed i q

(15)

As the grid-side voltages are balanced, eq = 0. eq = 0. The power factor is given to be unity for this project, the reactive power is zero, thus i q = 0. i q = 0. Therefore, the q-axis current reference is considered a constant of value ‘0 . The plant transfer functions for the d and q-axis of the grid current controller are G p (s) =

1 i d (s) = vd (s) Rg + s L g

(16)

G p (s) =

i q (s) 1 = vq (s) Rg + s L g

(17)

The value of L g . = 0.15 mH and.Rg .. s assumed to be zero. Figure 14 gives the block diagram of the current controller used in both part b and part c. The equations representing the d and q reference frame functions are: Lg

di d = ed − Rg i d − vd + ws L g i q dt

(18)

Lg

di q = eq − Rg i q − vq + ws L g i d dt

(19)

The PLECS schematic of the grid-side controller is shown in Fig. 15.

46

M. Rajendran and L. A. Kumar

Fig. 14 Block diagram of the current controller

Fig. 15 Grid-side current controller

The design of the dc voltage link controller and the grid-side current controller is done using the k-factor method in MATLAB and a summary of the control design parameters is given in Table 5.

Modeling and Simulation of a DFIG-Based Wind Energy System

47

Table 5 Design parameters for the DC link voltage and current controllers S. No.

Controller

ωc (Hz)

Kc

ωz (rad/s)

ωp (rad/s)

1

Current controller

2000

6346.9

3367.1

4689.8

2

DC voltage link controller

200

2115.6

336.7149

4689.8

4.3 Simulation Results of the Grid-Side Converter The entire model is simulated for two different wind velocities of 12 and 9 m/s and the corresponding waveforms along with the inference are discussed here. The waveforms shown in Figs. 16 and 17 are obtained by combining the 1.5 MW DFIG model, the wind turbine model, and the rotor-side converter that are connected to the 575 V, 60 Hz grid through the grid-side converter. Figures 16a, b, and 17a, b show the rotor speed and grid power waveforms at two different wind velocities of 12 and 9 m/s, respectively, when connected to the grid. It can be observed that the rotor speed and grid power reduce proportionally to the wind velocity. It can also be

Fig. 16 Waveforms for a wind velocity of 12 m/s a Rotor speed b Grid power c Rotor voltage d Stator voltage e Rotor currents f Grid currents

48

M. Rajendran and L. A. Kumar

Fig. 17 Waveforms for a wind velocity of 9 m/s a Rotor speed b Grid power c Rotor voltage d Stator voltage e Rotor currents f Grid currents

observed from Figs. 16d and 17d that the grid voltage is maintained at the desired value of 575 V irrespective of the changes in the wind velocity.

5 Conclusion The wind energy system based on the Doubly Fed Induction Generator (DFIG) was modelled using PLECS based on the mathematical expressions governing the machine operation. The system was analyzed for steady state and dynamic performance by perturbing the system with various wind velocities. The system was able to follow both step and ramp commands in the wind velocity. The outer speed loop and the inner current loop of the rotor-side converter (RSC) as well as the outer DC link voltage controller and the inner current controller of the grid-side converter (GSC) were designed using the k-factor approach of controller design. The system response was studied, and a comprehensive explanation of the obtained results was provided. It is also found that the values obtained from the analytical calculations and

Modeling and Simulation of a DFIG-Based Wind Energy System

49

the simulation have a close match and these analytical calculations are performed using MATLAB.

References 1. World Wind Energy Association Homepage, https://www.indea.org/information-2/inform ation/. Last Accessed on 18 Jan 2019 2. Indian Wind Energy Association Homepage, https://www.inwea.org/. Last Accessed on 18 Jan 2019 3. Arun Bhaskar M, Vidya B, Meenatchi V, Dhanduyathabani R (2011) DFIG based wind energy system (WES). In: Springer 2011 International conference on advances in communication network, and computing. Bangalore, India, pp 219–223 4. Sumathi S, Ashok Kumar L, Surekha P (2015) Solar PV and wind energy conversion system— an introduction to theory, modeling with MATLAB/SIMULINK and the role of soft computing techniques, 1st edn. Green Energy and Technology, Springer, New York, USA, pp 23–27 5. Bijaya P, Ojo O, Adeola B (2015) standalone operation of a DFIG-based wind turbine system with integrated energy storage. In: IEEE 2015 international symposium on power electronics for distributed generation systems. Aachen, Germany, pp 31–36 6. National Institute of Wind Energy Homepage, https://mnre.gov.in/national-institute-wind-ene rgy-niwe. Last Accessed on Jan 2019 7. Mukund RP (2006) Wind and solar power systems—design, analysis and operation, 2ndedn. Taylor and Francis Group, CRC Press, Boca Raton, FL, USA, pp 236–244 8. Akshay K (2013) DFIG-based wind power conversion system connected to grid. Int J Tech Res Appl 1:15–24 9. Rajeswari R (2014) Control of doubly Fed induction generator-based wind energy conversion system . Int J Adv Res Electri Electron Instrument Eng 3:381–389 10. PLEXIM DFIG Wind Turbine System Homepage, https://www.plexim.com/support/applic ation-examples/282. Last Accessed on 26 Jan 2018 11. Jackson JJ, Kyoung-Soo R (2012) Control strategies of doubly fed induction generator-based wind turbine system with new rotor current protection topology. J Renew Sustain Energy 4:423–432. 12. Tapia A, Tapia G, Ostolaza J (2003) Modeling and control of a wind turbine driven doubly fed induction generator. IEEE Trans Energy Convers 18:194–204 13. Raja A (2014) Personal communication. Arizona State University, Tempe, Arizona, USA 14. Meera GS, Divya NA (2015) Rotor side converter control of DFIG based wind energy conversion system. Int J Eng Res Technol (IJERT) 4:607–612 15. Peterson A, Herefords LT (2005) Evaluation of current control methods for wind turbines using doubly-fed induction machine. IEEE Trans Power Electron 20:227–235

Development of Wind Energy Technologies and Their Impact on Environment: A Review Manyamyuva Naga Satya Suryakiran, Waseemah Begum, R. S. Sudhakar, and Sharad Kumar Tiwari

Abstract Wind energy is a kind of renewable energy which produces electrical power from wind. Wind turbines are installed at a specific geographical location where there is abundant of wind. There are various environmental effects due to the operation and installation of the wind turbines that cannot be overlooked. In this paper, a status report of wind installation across the globe as well as environmental effects of installation of wind turbine has been discussed. Keywords Environmental impact · Global warming · Renewable energy · Wind energy

1 Introduction The way to choose sources of energy and decisions to impact Earth’s natural systems. So, it is necessary for us to choose the source of energy carefully. Wind energy is one of the sources of renewable energy, by which wind is used to generate electricity. It is derived from natural processes, i.e., either directly from the sun or from heat. The sources for electricity generation from solar, wind, hydropower, ocean, geothermal, biofuels, biomass, and hydrogen are derived from renewable sources [1]. The energy production from fossil fuels is also going to reduce because of the limited sources as per the demands. The International Energy Agency (IEA) estimated the worldwide energy demand will be increased by 1.6% yearly, approximately a 65% increase from developing countries. In REN21’s 2017 report, the contribution of renewable energy is 24.5% generation of electricity and 19.3% to energy consumption in the year 2015 and 2016. This energy consumption is coming 3.9% from hydroelectricity, 8.9% from traditional biomass, 4.2% from heat energy (biomass, solar heat and geothermal), and the residual 2.2% is electricity from solar, wind, geothermal, and various forms of M. N. S. Suryakiran · W. Begum · R. S. Sudhakar · S. K. Tiwari (B) Department of Electrical and Electronics Engineering, Aditya Engineering College, Surampalem, Andhra Pradesh, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_5

51

52

M. N. S. Suryakiran et al.

Fig. 1 World power generation (% of 24.973 TWh) (IEA 2018)

Natural Gas 23% Hydro 16%

Nuclear Fission 11%

Oil 4%

Non-Hydro Renewable 8%

Coal 38%

biomass. Approximately, US$ 288.9 billion amount is invested for the development of renewable technologies in 2018. There are some effects that can be combined with these resources. Wind power had a great history in mankind, i.e., more than 3000 years old. Humans started to practice it for the generation of electricity about 120 years back. The expansion of wind was always dependent on the oil prices that make the wind power to get the chance to show their power during the 1970s [2]. During the last decade, there is plenty of research and development going on wind turbine design and its development was carried in this period. As compared to the pioneering countries, such as USA, Germany, Spain, and Denmark have made substantial effort to progress their wind power industry [2–4]. It is expected that wind energy system will provide 5% of the world’s energy in 2020 [5] (Fig. 1). In the last decade, it performed well and also makes a better environmental impact, i.e., climatic and visual impact, noise and pollution, etc. But it produces very minor energy production as compared to fossil fuels. Wind can be an effective power source and a high potential to accomplish energy needs. The usage of wind energy has only a partial influence on the environment. The aim of this paper is to offer an overview of wind energy framework, the progress trend of offshore wind power, the current development of wind turbines, and the environmental and climatic impact on living beings. It also presents various analyses for wind studies, its development, and the impacts on the environment.

2 Historical Background Wind energy is not new to humans as humans began to use it about 3000 years ago in Egypt. They initiated to construct a wind mill for pump water. China also uses the wind wheels with a vertical axis of rotation to drain rice fields, but centuries before, European started to used it and also the horizontal windmill was developed in Europe, in 1180, in Duchy of Normandy [6].

Development of Wind Energy Technologies and Their Impact …

53

Wind energy in USA had been developed since 1850 by establishing a wind engine company. In 1890, there is a steel blade introduced to the windmills and wind power is used for pumping water and electricity. And during World War II also, there is a utility of wind energy to electric power. In 1970, due to the increase in oil prices, there is a boost up to the wind turbine inventions. During 1980, there is a construction of the first large wind farm [7, 8]. Wind energy in India began in 1986 where the original wind form was built in coastal areas of Tamil Nadu, Gujarat, and Maharashtra with 55 Kw Vestas wind turbines. India has the world’s fourth largest installed wind power [9].

3 Current Status of World Wind Energy Total worldwide capacity of installed wind power plant till the year 2018 for electricity generation is 591,549 MW, and it has been increasing by 9.6% compared to the previous year [10]. Installations are increased by 54,642 MW, 63,330 MW, 51,675 MW, and 36,023 MW in 2016, 2015, 2014, 2013, respectively. Some countries have reached comparatively high levels of wind power penetration, such as 39% of stationary electricity production in Denmark [11, 12], 9% in Germany, 16% in Spain [13], 14% in Ireland [14], and 18% in Portugal [13] in 2010 [13, 14]. Eighty-three countries are using energy on a commercial basis as of 2011 [15]. 3.1% wind power’s share in worldwide at the end of 2014 [16]. Scotland crossed the threshold of wind energy supply 100% of electricity needs for the country in November 2018 [17].

3.1 Installed Wind Power Capacity in Particular Countries Installed wind power capacities by country are provided in this section; the data is obtained from the global wind energy council. The worldwide wind power capacity is increased by 63,330 MW or 17.14% from 369,553 MW to 43,283 MW in 2015 [17].

3.1.1

China

China is a late arrival in wind energy, but it has caught up fast. Now, China has become the world’s largest adopter of wind power, comprising 33% of global wind power capacity (BP 2016), far ahead of the USA (17%) and Germany (10%) (Table 1). China does not develop wind energy in isolation but has learned much from international peers in both technology and policy arenas [18]. In [19] argue that China’s

54

M. N. S. Suryakiran et al.

innovation system for wind energy involves a process of system-building that differentiates it from those of countries where wind technology originates. Reference [20] shows that China has acquired advanced wind technology by building access to technical know-how that originated in developed countries through international technology transfer and mergers and acquisitions (M&A). Reference [21] demonstrate how the Chinese wind turbine manufacturer Envision Energy has upgraded innovation capability by tapping into the Danish innovation system. GWEC and IRENA (2012) argue that China’s rapid development of wind power attributes to a strong long-term legislative background, a clear tariff structure, and a strong industrial base.

3.1.2

USA

More than 50,000 wind turbines are placed across USA. This produces about 8% of the nation’s energy generating capacity and they estimate that the percentage will reach 20% by 2030 [22]. As per the Lawrence Berkeley Laboratory report, turbines at wind farms are getting bigger and more efficient. It has derived 2.32 MW average turbine capacity and up 8% compared to the previous year and which is greater than 200% since the late 1990s. Blades, rotors, and turbines are getting bigger with more projects and exceeding levels, which has been required special permits from federal aviation regulators. Turbines constructed between 2014 and 2016 had an average “capacity factor” of 42%, related to an average capacity factor of 31.5% for projects built from 2004 to 2011 [23]. The Vineyard Wind project in Massachusetts may be in progress coming year and accomplished in 2021. It is possible to become the country’s largest offshore wind farm at 800 MW. Various states in USA have offshore wind projects that are available for leasing, i.e., Delaware, New Jersey, California, Maryland, New York Connecticut, Virginia, Hawaii, Massachusetts, Maine, North Carolina, Rhode Island, and Ohio. Offshore wind is an emerging possibility since it is near to the major population centers while for the onshore or solar system, it would be difficult in largely populated areas. Though with planned projects, USA is too behind in the offshore wind system. The Germany, UK, and China have 4,667, 5824 MW, and 1823 MW, respectively.

3.1.3

Japan

The wind power generation in Japan is an insignificant percentage of the country’s electricity. The target of the Japanese government till 2030 for wind power generation is relatively low when compared to other countries, at 1.7% of electricity generation. C-tech’s wind farm (Shin Aoyama Wind Farm) is the largest wind farm in Japan, as of 2018. It has 40 turbines with a capacity of 80 MW and the amount of electrical energy that can be produced are approximate to the yearly consumption of about 44,000 conventional households. The total installed capacity is 3,399 MW. It is expected that the Yurihonjo Offshore Wind Farm project may begin in 2021. It may include

Development of Wind Energy Technologies and Their Impact …

55

70–90 turbines with a capacity of 1000 MW. It also becomes the largest offshore wind farm in Japan. In 2013, the Japanese government tested a project based on floating offshore wind turbines for 1 km off the coast at Kabajima in Nagasaki Prefecture. It has 608 GW of offshore and 144 GW for onshore wind capacity [24].

3.1.4

Spain and Germany

Spain and Germany are the two top wind power producers in Europe. By 2018, installed capacity of onshore and offshore wind power had been reached to 88 GW and 7.6 GW by 2020, respectively [10]. The Energiewende is proposed by German in late 2010 to low their carbon rates and the greenhouse gases drops of 80–95% by 2050 and to increase renewable energy to 60%c by 2050. By 2016, the citizens in Germany also supporting the Energiewende that about 80–90% of the public is in favor. One reason for the high acceptance rate in Germany is because of participation in the Energiewende, as private, landowners, households, or members of energy cooperatives [11] (Fig. 2). Spain covers 19% of electricity consumption and having 1123 wind farms present in 807 municipalities, with 23,308 installed wind turbines, which [12]. It is estimated as in Spain, wind turbines have been contributing 19% of the electricity spent in 2018. Spain is the second European country that installed wind farms and fifth worldwide. Last year, it was increased by 392 MW, which makes the total installed capacity in Spain, in December 2018, at 23,484 MW, as per data collected by the Spanish Wind Energy Association. Fig. 2 German electricity by source in 2017 [9]

Others 13% Brown Coal : 134 Twh 24%

Wind: 103.6 TWh

Solar: 38.4 Natural TWh Gas: 49.1 7% TWh 9% Nuclear: 72.2 TWh 13%

Hard Coal: 81.7 TWh 15%

56

3.1.5

M. N. S. Suryakiran et al.

India

Wind power in India had significantly emerged from the last decade. India becomes fourth-largest installed capacity, i.e., 36.625 GW. The installations in India are placed in East, West, and South regions and also reduce the cost.

4 Advantages of Offshore Wind Wind on offshore is more powerful than land so to obtain more power to wind turbine, ocean wind is better than the land wind and due to this reason, the wind turbines are placed in the offshore with more than 20 Km far from land. At that place, the wind flow is more so placing turbine at that place is more effective and can generate more power. The characteristics of the layer of turbulent air adjacent to the ground and sea surface allow the offshore turbine [12]. The second benefit is that there is large area to install large number of wind turbines and help to produce greater power from more turbines. Due to far away from the sea shore, the noise produced by them will not affect the humans and they can rotate with more speed. Placing them far away also reduce the visionary disturbance [12, 13] (Figs. 3 and 4). The wind turbine is a domestic energy source and creates new jobs in world and also eco-friendly by not emitting the green-house gases and other harmful pollutants [13]. Due to these more advantages to offshore, wind energy every country in world is been looking on it, and placing the money in this renewable energy construction is more useful to the long seashore countries that gives them better electric power generation and economical rather than small seashore because the cost of the construction is two times more than onshore wind turbines. Fig. 3 March 2018 [8] total capacity (MW)

2,520

101

4

53 8,197

3,963 4,298

5,613 4,509 4,784

Tamil Nadu Karnataka Madhya Pradesh Others

Gujarat Rajasthan Telangana

Maharashtra Andhra Pradesh Kerala

Development of Wind Energy Technologies and Their Impact … Fig. 4 Comparative analysis between the potential of various offshore and onshore energy demand and depths in Europe [12]

57

EU Demand 2020 EU Demand 2010 EU Demand 2000 Potenal Onshore 40m Potenal Onshore 30m Potenal Onshore 20m Potenal Onshore 10m Potenal Onshore 0

1000

2000

3000

4000

5 Wind Energy Induced Environmental Issues As the world is searching for renewable energies due to the effect of greenhouse gases produced by fossil fuel emission. So there is a great demand for renewable resource technologies. The new technologies also have a somewhat effect on the earth’s environment but there is a minor effect on environment rather than fossil fuels. Technology needs to locate these problems quickly and solve them for better usage. One of the technologies that may cause small effect on the environment is wind energy, and wind turbines had become the latest trend in the world for producing power from wind. So, a number of wind turbines are installed in many countries for producing more power. But it also causes the minor impact on environment. A wind power plant converts wind energy into electrical energy using wind turbines. The output of turbine depends on area swept by the turbine blades, density of air, and wind speed [16]. The major issues related to wind turbine on environment are bio-system disturbance, visual pollution, wildlife safety, electromagnetic interference, noise, and local climate change [17, 18].

5.1 Birds Wind turbines are causing a great difficulty to the bird species, due to placing the wind turbines at the high altitude and wide range areas for the maximum power production. But this may cause the affect on migration birds and birds living on high altitudes. While moving birds from one place to other, they are colliding with the wind turbines, and by this reason, the bird mortality rate is increasing year by year in all countries [19]. The normal height of wind turbines is about 328 feet from ground [22]. Specific type of species (75 golden eagles die every year due to the farm’s 5000 turbines) killed by wind turbines is a major concern [20].

58

M. N. S. Suryakiran et al.

Due to collision of birds with the turbine, it is estimated that 140,000–328,000 birds are died every [21]. It also creates wind disturbance in the area which causes disturbance in migration paths. Therefore, various birds are flying away from the wind turbine area. Therefore, birds require extra energy to travel; some birds like eagles and vultures travel along speed and also at high altitudes but placing turbines makes the bird mortality to increase in a great manner.[23]. The diameter of wind turbine is about 40–90 m [24] and this also causes the major risk at night time for birds because the birds cannot alter their direction in a short period because of long diameter and they collide [25].

5.2 Bats Bats are special creatures in the earth that they have different features for hunting its prey; they always use to hunt their food during night rather than day time. They use its hearing to navigate at night and use echolocation to catch insects continuously. They emit ultrasonic waves with high frequencies (20–100 kHz). The waves are reflected by hitting the various objects and echoes have been returned to bats. The echo signals enable the bats to form a mental map of its environment [25]. The birds use the wind currents to fly but the wind turbines cause changes in wind current at that particular area around the turbine, and so these disturbances cause the bats to fall in to the blades of turbine. According to the night study and condition, delaying the cut-in speed reduces the bat deaths by 44–93%. The flying animals run into the spinning blades, or the quick decrease in the air pressure around the plant can cause barotrauma (lungs bleeded) [26]. Bats use the eco-location to locate the prey but many preys are located at the wind turbine due to heat and light sources and because of it, the bats are attracted to the turbines and fall to ground. The main reasons that bats are attracted to wind turbines are auditory attraction, electromagnetic field dis-orientation, insect and heat attraction, forest edge effect, and thermal inversion [27].

5.3 Marine Species The marine species are affected majorly due to offshore wind turbines because they were placed at the 20–40 km away from the sea shore. The marine life is not more at the coastal rather than the in-depth ocean, while installing the turbine, they are drilling about >60 m, which will affect the underwater species [28]. The offshore wind farms are increasing as they have high potential to generate high power and constant energy due to continuous flow of air. Approximately, 3,230 turbines at 84 offshore farms were installed across 11 European countries having total capacity of 11,027 MW [29, 30]. Increasing the greater number of turbines in ocean can cause disturbance to the marine life. The wind turbines produce the noise while they rotate,

Development of Wind Energy Technologies and Their Impact …

59

but in ocean, it may be much high due to continuous high flow of air, and this may result in the loud sounds [31]. It is observed that in underwater turbine, foundations regularly transform into artificial reefs, attracting small fish, and mollusks that feed on plankton. This may create disturbance in the food chain to seals, dolphins, and larger fish [32, 33].

5.4 Deforestation and Soil Erosion While placing the wind turbine farm it required a wide range of areas and the area should be in the wide areas of forest land and hill regions because the high speed of airflow is present at that places. Most of the places suitable for wind farms are located in the forest lands. The basic concept for wind farm spacing is that the rotor diameters of each blade are away 7 m from each other [34]. Because of this, the land requirement is more and to install the wind farm, they required plane areas of the forest so there is a huge amount of deforestation that will occur in forest areas, and this affects the habitat present in that area. Due to deforestation, there is also a change in climate conditions in particular areas. The base of the tower is fixed in a platform of steel rebar and more than 1000 tons of concrete at 30–50 feet across and anywhere from 6 to 30 feet deep [35]. While doing the installation, they dig up to 30 feet and also pour the cement into it, and by this, there is a cause of soil disturbance and it will cause soil erosion surrounding that area. When the windmill needs to repair or dismantle after its lifetime, then the process of removing it is very hard due to the huge concrete installation at the beginning and causes a huge impact on the soil while removing the windmill after its life-time.

5.5 Noise Noise is a disturbance or unwanted sounds, which our ears cannot be tolerated. Exposure to prolong or excessive noise can cause a high range of health problems such as stress, loss of concentration, hearing impact, sleep disturbance, and heart diseases [36]. Wind turbine causes the sound approximately 43 decibels and it will be reduced based on distance and at 500 m, sound is about 38 decibels and beyond the mile, we cannot hear the noise [37]. The human normal audible range is 40 decibels and above of it causes the hearing damage to humans [38]. Wind turbine noise not only disturb the human but also aquatic animals due to offshore wind farm; the fishes have less audible range than humans so the noise from turbine causes high damage to them [39]. The rotor blades may produce aerodynamic noise in oceans through an airborne path. As wind speed increases, sound levels also increase slightly. The kind of wind turbine basis will also affect the transmission of the under-water sound [40]. Noise levels in oceans can create disturbance to the dolphins [41].

60

M. N. S. Suryakiran et al.

5.6 Visual Impact The wind turbines are eco-friendly and there is very less impact on the environment, but due to their huge structures, they are causing the visual disturbance to the natural views and human eyes. Due to rotating turbine blades, sunlight producing unavoidable flicker bright is also interrupted known as shadow flicker [42]. The shadow flicker induces adverse effects on humans [43]. WHO found that the risk of depression is increased by 40%, the bad view of window [44]. On a special report by the International Board on Climate Change summarizes wind energy and climate change mitigation [45].

5.7 Reception of Radio Waves and Weather Radar The wind turbines emit the electromagnetic radiation as well as heat while they rotate, and this emission may affect the signal waves of transmission. .The transmission waves will pass from high altitudes from ground, but these turbines are also placed at large heights; this causes the interaction of transmission signals by the radiation of wind turbines. If the transmitter antenna is near to the turbine, it has two effects. First is diffraction due to pylon, and the second is the reflection [46]. It also causes the interference in radar systems, because of placing turbines on the opposite side of elevated terrain. So that the turbines are blocked from view of the radar; relocating planned turbines that might pose interference issues outside of the radar line of sight; lining turbines up along radials 3 from the radar to minimize the azimuthal area impacted by the turbines; or spacing the specific locations of wind turbines farther apart to enable detection of targets between them [47]. Through joint activities and investments, the Wind Turbine Radar Interference Mitigation group try to remove radar interference as an impediment to future wind energy development [48].

6 Conclusion Renewable energy, also known as sustainable energy or green energy, is a source which can be replenished. It is one solution for the global energy problem. The need of renewable sources continues to increase as fossil fuel depletes. This energy has been useful in socioeconomic impacts such as rural development openings, making local industry, diversifying the energy supply, and employment chances. However, it can cause environmental problems in a habitat. Trivial issues may origin tragic effects in the future when wind energy becomes one of the major energy sources. As mentioned in this study, more systematic studies are required on the potential impacts of wind farms on the environment. Wind energy mistreatment should be assessed for the ecological influences, environmental, economic, and social platform.

Development of Wind Energy Technologies and Their Impact … Table 1 Global wind power capacity (BP 2016)

Countries

% Power capacity

China

33

USA

17

Germany

10

61

References 1. Wikipedia: https://en.wikipedia.org/wiki/Wind_power 2. Ackermann T, Söder L (2002) An overview of wind energy-status 2002. Renew Sustain Energy Rev 6(1–2):67–127 3. Arslan O (2010) Technoeconomic analysis of electricity generation from wind energy in Kutahya. Turkey Energy 35(1):120–131 4. Hepbasli A, Ozgener O (2004) A review on the development of wind energy in Turkey. Renew Sustain Energy Rev 8(3):257–276 5. Xu J, He D, Zhao X (2010) Status and prospects of Chinese wind energy. Energy 35(11):4439– 4444 6. Herbert GJ, Iniyan S, Sreevalsan E, Rajapandian S (2007) A review of wind energy technologies. Renew Sustain Energy Rev 11(6):1117–1145 7. https://www.energy.gov/eere/wind/history-us-wind-energy 8. https://atomberg.com/wind-power-in-india/ 9. https://en.wikipedia.org/wiki/Energiewende_in_Germany 10. Snyder B, Kaiser MJ (2009) Ecological and economic cost-benefit analysis of offshore wind energy. Renewable Energy 34(6):1567–1578 11. https://www.evwind.es/2019/02/20/spain-increased-wind-power-by-392-mw-in-2018-add ing-a-total-of-23484-mw-installed-wind-energy/66205 12. https://www.americangeosciences.org/critical-issues/faq/what-are-advantages-and-disadvant ages-offshore-wind-farms 13. Knopper LD, Ollson CA (2011) Health effects and wind turbines: a review of the literature. Environm health 10(1):78 14. Lima F, Ferreira P, Vieira F (2013) Strategic impact management of wind power projects. Renew Sustain Energy Rev 25:277–290 15. Premalatha M, Abbasi T, Abbasi SA (2014) Wind energy: Increasing deployment, rising environmental concerns. Renew Sustain Energy Rev 31:270–288 16. https://large.stanford.edu/courses/2016/ph240/white1/ 17. Kingsley A, Whittam B (2001) Potential impacts of wind turbines on birds at North Cape, Prince Edward Island. Bird Studies Canada, Sacksville, New Brunswick 18. https://www.nbcnews.com/mach/science/wind-energy-takes-toll-birds-now-there-s-help-ncn a866336 19. https://www.wind-watch.org/faq-size.php 20. https://www.carbonbrief.org/bird-death-and-wind-turbines-a-look-at-the-evidence 21. https://en.wikipedia.org/wiki/Wind_turbine_design 22. Dai K, Bergot A, Liang C, Xiang WN, Huang Z (2015) Environmental issues associated with wind energy–a review. Renew Energy 75:911–921 23. Kerr RA (2007) Global warming is changing the world. Science 316(5822):188–190 24. https://www.sciencedaily.com/releases/2015/06/150619084612.htm 25. https://www.scientificamerican.com/article/bat-killings-by-wind-energy-turbines-continue/ 26. https://electrical-engineering-portal.com/why-bats-are-insanely-attracted-to-wind-turbines 27. https://www.ewea.org/statistics/european/ 28. Ho A, Mbistrova A (2016) & Corbetta, GThe European offshore wind industry-key trends and statistics 2015. Belgium, Brussels

62

M. N. S. Suryakiran et al.

29. 30. 31. 32.

https://www.sciencedaily.com/releases/2014/10/141016123608.htm https://www.dw.com/en/how-do-offshore-wind-farms-affect-ocean-ecosystems/a-40969339 https://iopscience.iop.org/article/10.1088/1748-9326/9/3/034012/meta Gill AB (2005) Offshore renewable energy: ecological implications of generating electricity in the coastal zone. J Appl Ecol 42(4):605–615 Westerberg H, Lagenfelt I (2008) Sub-sea power cables and the migration behaviour of the European eel. Fish Manage Ecol 15(5–6):369–375 https://sciencing.com/much-land-needed-wind-turbines-12304634.html https://www.science.org.au/curious/earth-environment/health-effects environmental-noisepollution https://www.ge.com/reports/post/92442325225/how-loud-is-a-wind-turbine/ https://www.cochlea.org/en/hear/human-auditory-range https://en.wikipedia.org/wiki/Hearing_range https://dosits.org/animals/effects-ofsound/anthropogenic-sources/wind-turbine/ https://www.sciencedaily.com/releases/2018/10/181024112225.htm Harding G, Harding P, Wilkins A (2008) Wind turbines, flicker, and photosensitive epilepsy: characterizing the flashing that may precipitate seizures and optimizing guidelines to prevent them. Epilepsia 49(6):1095–1098 https://www.energy.gov/sites/prod/files/2015/03/f20/wv_chapter2_wind_power_in_the_uni ted_states.pdf Salema C, Fernandes C, Fauro L (2001) TV interference from wind turbines. Proc Conf Telecommun ConfTele (1):47–51 https://www.energy.gov/sites/prod/files/2016/06/f32/Federal-Interagency-Wind-TurbineRadar-Interference-Mitigation-Strategy-02092016rev.pdf Bolinger MW (2010) Wind technologies market report. No. DOE/GO-102011–3322. National Renewable Energy Lab.(NREL), Golden, CO (United States) https://www.energy.gov/sites/prod/files/2019/04/f61/WTRM_Factsheet_Final_2019.pdf GWEC, Global Wind Report Annual Market Update Gwec.net. Retrieved 20 May 2017. World Wind Energy Report 2010 (PDF). Report. World Wind Energy Association. February 2011. Retrieved 30 April 2011.

33. 34. 35. 36. 37. 38. 39. 40. 41.

42. 43. 44. 45. 46. 47. 48.

Regulated Jordan—Elman Neural Network-Based Controller Model for Grid-Connected Wind Energy Conversion Systems S. N. Deepa and N. Rajasingam

Abstract Jordan—Elman neural network (JENN) is a class of recurrent neural network that utilizes their internal unit to process pattern of samples. JENN simply extends the multilayer perceptron by enrolling context neurons which are those internal units. The main focus of this paper is to design and develop an effective controller model for doubly fed induction generator (DFIG)-based grid-connected wind energy systems. To improve the performance of considered grid-connected wind energy conversion system (WECS), a meta-heuristic optimization algorithm namely density-based gray artificial bee colony (D-GABC) algorithm is utilized to effectively train the JENN model. Once the optimized neural network architecture is modelled, it is used to adjust the gain values of the traditional PID controller so as to maximize the energy harvesting. Conducted numerical simulation results prove the effectiveness of the developed regulated Jordan—Elman neural network (RJENN) than that of methods available in literature. Keywords PID controller · Jordan—Elman neural network · Artificial bee colony algorithm · Doubly fed induction generator · Wind energy conversion systems

1 Introduction Neural networks are widely associated as an effective strategy for handling nonlinear and dynamic samplings, specifically in the conditions where the underlying relationships do not entirely presume. Neural architectural model affords a space to enact computation with a vast amount of samples due to their parallel functioning. However, with regard to refine the network mechanism and attain better detection and potency, additional efforts are required [1, 2]. S. N. Deepa (B) · N. Rajasingam Department of Electrical and Electronics Engineering, Anna University Regional Campus Coimbatore, Tamil Nadu, Coimbatore 641046, India e-mail: [email protected] N. Rajasingam e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_6

63

64

S. N. Deepa and N. Rajasingam

Recurrent neural networks are a class of neural networks which are widely applied for several sequential applications such as machine translation, token-level classification, biology computations, captioning of images, speech recognition, and sentence classification. The recurrent network is comparable with that of the feed-forward network, where it gets diverged only in feedback mechanism, i.e., the hidden layer response is fed back into the input layer. In other words, these networks are typically back-propagation neural networks with appropriate feedback links [3]. One of the popularized classes of the recurrent neural network is the Jordan— Elman neural network (JENN). This neural net is formerly proposed by Jordan and thereupon remodelled by Elman [4, 5]. For the purpose of averting the gradient problems, the hidden nodes of this network should bear the training associated with parameter update process. In this work, the optimal learning of JENN is accomplished by employing the swarm intelligence. Density-based GABC strategy is applied into JENN for the purpose of further enhancement in network performance and on account of evolution of DFIG controller model to determine optimal design. A dynamic modelling and control of DFIG was illustrated for wind turbine systems. In consideration of obtaining a better power flow transfer and to enhance the transient stability of dynamic system, an improved feedback control method was adopted. The feedback mechanism handles the active and reactive power control through enacting the appropriate voltage vectors on the rotor-side [6]. A new robust fractional order sliding mode controller was conferred for maximum power point tracking control of DFIG-based WECS. Disturbances and uncertainties are evaluated using a chattering-free fractional order uncertainty estimator for the purpose of improving the stability of the control system [7]. A new grouped gray wolf optimizer (GGWO) was developed to optimize the interactive PI controller parameters of DFIG-based wind turbine. GGWO was molded by introducing a ‘grouping’ mechanism to typical meta-heuristic algorithm named gray wolf optimizer. Under the constructed framework, maximum power point tracking was attained with an enhanced fault ride through capability [8]. A simplified modelling method was also developed by analyzing the control strategy and network topology of DFIG with the stator flux-oriented vector control technique. Further, in both grid-side and rotor-side of DFIG, energy balance-based controlled voltage source was deployed as alternative to the pulse width modulation control [9]. An intelligent control system was evolved by applying model-free control polices for DFIG-based wind turbines. The online control system works on an adaptive actor-critic technique-based policy iteration reinforcement learning model [10]. A coordinated high order sliding mode control mechanism was also presented for utility grid synchronization and power optimization in a DFIG-based WECS. The acquired wind energy was maximized with optimal rotor speed tracking by applying supertwisting algorithm; meanwhile the stator voltage is synchronized with the utility grid through synchronization technique, without using the current control loop [11]. The above-detailed review of literature clearly demonstrates that certain robust techniques are required to overcome the disadvantageous influence of variations in the system control parameters and to design an expert system model. In turn, a hybrid D-GABC-JENN architecture is designed in this paper for DFIG driven by

Regulated Jordan—Elman Neural Network-Based Controller Model …

65

variable speed wind turbine in order to optimize the wind power extraction in the grid-connected wind energy systems.

2 Grid-Connected Wind Energy Systems The wind energy systems transform the kinetic energy present in the wind to electrical energy. Presently, majority wind energy systems preferably adopt the doubly fed induction generator (DFIG) due to its superior merits including flexibility of active and reactive power control, relatively low mechanical stress on the wind turbine, reduced flicker, maximum power generation under wind speed variations and low investment. Also, the rotor speed of DFIG can be regulated from sub-synchronous mode to super-synchronous mode [12]. Figure 1 depicts the schematic representation of standard DFIG-based wind energy conversion system. The complete mathematical modelling of DFIG is available in Ref. [13]. The standard form of synchronously rotating dq-coordinate equations in per unit (pu) system is taken into consideration for modelling a DFIG. In spite of the steady-state considerations along reference frames, the mathematical model of the considered DFIG system in the aspect of transfer function form is formulated to employ in the optimal controller design and given as follows [14, 15]: G(S) =

0.000324s 6 − 1.75s 5 − 2366s 4 + 7.9e6 s 3 + 7.5e9 s 2 = 5e12 s + 2018e14 s 6 = 2340s 6 + 8.67e6 s 5 + 4.7e9 s 3 + 2.7e12 s 2 + a1.27e14 s + 9.6e14 (1)

This derived transfer function is utilized to perform the regular PID controller design and as well as in proposed neural network-based controller modelling for DFIG-based wind energy systems. The performance criterion to tune the developed D-GABC-based JENN controller model is described as, JI = (1 − e−κ )(Povershoot + esteady_state ) + e−κ (Tsettling − Trise )

Fig. 1 Schematic representation of DFIG

(2)

66

S. N. Deepa and N. Rajasingam

where ‘κ’ is the scaling factor, ‘Povershoot ’ is the peak overshoot, ‘esteady_state ’ is the steady-state error, ‘T settling ’ is the settling time, and ‘T rise ’ is the rise time. This derived Eq. (2) is perceived to be the performance criteria with regard to regulate the controller gain parameters for sixth-order transfer function form of DFIG mathematical model evolved in Eq. (1).

3 Regulated Jordan—Elman Neural Network 3.1 Jordan—Elman Neural Network Jordan—Elman neural network is a simplified form of a recurrent neural net, where the multilayer perceptron is extended with context neurons that are the processing elements (PEs) in the neural mechanism. The context unit remembers the past activity by means of current gradient values where it forgets the past input with an exponential decay. This signifies the incident that happened is forceful than the ones that have occurred further in the past. The context units confer the ability to extract temporal information from the data to the neural network. The context unit leads the forgetting factor by means of the time constant. The input sample data pattern is characterized through three stages namely learning, cross-validation, and testing [4, 5, 12]. Figure 2 shows the architecture of the standard JENN model.

Fig. 2 Architecture of the Jordan—Elman neural network

Regulated Jordan—Elman Neural Network-Based Controller Model …

67

The input layer has ‘m’ neurons as independent variables, while the ‘o’ neurons in the output layer are the dependent variables. The interaction between the input parameters is represented by the hidden layer. Initially, the network’s output is computed and then compared with a desired outcome, resulting in an error strands for each output neurons. Subsequently, the estimated error strand is applied to each neuron in the network, and the corresponding weights are refined. In Jordan network, the context units are enacted from the output layer, whereas the PEs are enacted from the hidden layer in case of Elman network. The JENN model output gets extracted from combination of Jordan and Elman nets. For the purpose of enhancement, PEs are fed from both the hidden and the output layers. Later, the context layer subjoins to the input layer as regard to thoroughly interconnecting with the hidden layer. PEs store the activations of the output neurons from the previous time step by means of the feedback links with weights. Thus, PE performs as a low pass filter that produces the response as a weighted value of recently occurred outcomes. Subsequently, the network output is acquired through summing the recently occurred outcomes multiplied with a scalar parameter. The input vector of Jordan network is described as, J X = x1 , x2 , ..., xm , xm+1 , xm+2 , ..., xm+1+o       input units

(3)

output context units

and their corresponding output of each neuron in the output layer is calculated as; ⎛ m+1+o

⎞ n+1   Yk = f yk ⎝ w jk f z j wi j xi ⎠ j=1

(4)

i=1

where (xm+2 , ..., xm+1+o ) = (y1 (t − 1), y2 (t − 1), ..., yk (t − 1))

(5)

‘x’ is the input training vector, ‘y’ is the actual output vector, ‘wij ’ is the weight values of interconnection between the ith input unit and the jth hidden unit, and ‘wjk ’ is the weight values of interconnection between the jth hidden unit and the kth output unit. The input vector of Elman network is described as, X = x1 , x2 , ..., xm , xm+1 , xm+2 , ..., xm+1+n       input units

(6)

hidden context units

and their corresponding output of each neuron in the output layer is calculated as; ⎛ Yk = f yk ⎝

n+1  j=1

w jk f z j

m+1+n  i=1

⎞ wi j xi ⎠

(7)

68

S. N. Deepa and N. Rajasingam

where

 (xm+2 , ..., xm+1+n ) = z 1 (t − 1), z 2 (t − 1), ..., z j (t − 1)

(8)

and ‘z’ is the hidden vector. JENN is configured by incorporating a Jordan neural net with an Elman neural network. Hence, it combines the past values of PEs with the present input data in order to obtain the net output. Output vector of JENN is formulated by combining Eqs. (4) and (7). The input vector of Jordan—Elman network is given by, X = x1 , x2 , ..., xm , xm+1 , xm+2 , ..., xm+1+n , xm+2 , ..., xm+1+o          input units

hidden context units

(9)

output context units

and their corresponding output of each neuron in the output layer is calculated as; ⎛ m+1+n+o

⎞ n+1   Yk = f yk ⎝ w jk f z j wi j xi ⎠ j=1

(10)

i=1

where  (xm+2 , ..., xm+1+n , xm+2 ..., xm+1+o ) =

z 1 (t − 1), z 2 (t − 1), ..., z j (t − 1), y1 (t − 1), y2 (t − 1), ..., yk (t − 1)

 (11)

The training and generalization ability of the substantiated JENN model is estimated with respect to the performance measure such as MSE. The error criteria for refining the neural net is described as, P MSE =

j=0

2 N i=0 di j − yi j N·P

(12)

where ‘P’ is the number of output PEs, ‘N’ is the number of exemplars in the data set, ‘y’ is the actual output vector, and ‘d’ is the desired output vector. However, this regular JENN while employed in DFIG-based wind energy systems experiences some common issues such as trapped into local optimum and tends to premature convergence and stagnation problem [13]. Hence, this paper endeavors to conquer the drawbacks by integrating the density-based GABC algorithm with the usual JENN model.

Regulated Jordan—Elman Neural Network-Based Controller Model …

69

3.2 Density-Based Gray Artificial Bee Colony Algorithm To achieve a relatively best neighbor in traditional artificial bee colony algorithm (ABC) that originally developed by Karaboga in 2005 [16], a gray artificial bee colony (GABC) algorithm was proposed by Xiang et al., where the gray relational degree was introduced between a present individual and its neighbors [17]. The applicability of GABC theory on controller modelling for wind energy conversion system resulted in slow convergence rate and also attains premature convergence of the network model. This initiated the requirement of perfect balance between the exploration and exploitation behavior of bees in the search domain, where it is achieved by introduction of probability density function in GABC algorithm [15]. On artificial employed bees phase, the combinatorial candidate solution equation is given by,

yi, j

 

⎧ ⎨ xk, j + φi, j ∗ xi, j − xk, j +  φi, j ∗ x dence_best − xk, j  if φi, j < α = xbest, j + φi, j ∗ xi, j − xbest, j + φi, j ∗ xdence_best − xk, j if φi, j < μ ⎩ ub x lb j + x j − x i, j otherwise (13)

where ‘α’ and ‘μ’ are predefined gray relational degree-based probability values of corresponding solution equations, respectively. The combinatorial solution search equation for the artificial scout bees phase is written as,    ub lb x lb if φi, j < ε j + ϕ ∗ xj − xj . (14) x˜i, j = lb ub x j + x j − xi, j otherwise where ‘ε’ is the predefined probability value, ‘φ’ is a random real number in the range of [0,1], ‘ϕ i,j ’ is a random real number between [−1,1], ‘x j lb ’ is the lower bound of jth index, ‘x j ub ’ is the upper bound of jth index, ‘best’ represents the index of the best individual found so far, and ‘x dense_best ’ is the local maxima point of all the peaks of the considered data points obtained employing the density estimate procedure for normal density function and Rayleigh density function that is associated with a novel bandwidth estimation method for addressing the premature convergence. Based on this dense position point, the solutions are calculated with respect to the movement of the bees in the solution search equation, and the search process is enhanced to balance exploration and exploitation achieving a better optimal solution. The detailed pseudo code for density-based GABC algorithm is available in Ref. [15].

3.3 Regulated Jordan—Elman Neural Network A major limitation in an artificial neural network-based controller modelling is the generalization capability or robustness of the designed structures, i.e., how

70

S. N. Deepa and N. Rajasingam

sustainably the developed controller model works on unpredicted data. Hence, JENN is propelled to involve an optimization scheme based on density-based GABC algorithm with respect to make out their optimal operating provisions which lead to a maximum performance within a minimum response time. The density-based GABC algorithm is hybridized to generate the optimal interconnection weight values for JENN. In the new developed JENN, the weights of all the input layer, hidden layer, context recurrent layer, and output layer are tuned for best values using the D-GABC algorithm, and also in the new regulated JENN, the weight correction carried out between the hidden layer and context layer includes the momentum factor correction. This inclusion of momentum factor correction achieves the existence of recurrent layer more prominent and enhances the network performance. Density-based GABC algorithm is invoked when the interconnection weights do not get optimized over maximum number of iterations in the new R-JENN model. The introduction of density estimate enhances the exploitation rate of the densest source search process. The search process for ideal densest source in the designed density-based GABC algorithm is conducted in a way to obtain the optimal dataset for conventional JENN. Later on, the developed optimized Jordan—Elman neural net, named as regulated Jordan-Elman neural network (R-JENN), is utilized to regulate the PID controller gain parameters. Figure 3 shows the structure of proposed regulated JENN model. The algorithmic procedure of DFIG system description for D-GABC-JENN-PID controller model is described in the following sub-section.

Fig. 3 Structure of proposed R-JENN controller model

Regulated Jordan—Elman Neural Network-Based Controller Model …

71

4 Learning Algorithm of Proposed Regulated JENN Model Thus, R-JENN is trained until the network converged or meets the error criterion or reaches a certain number of iterations. On the basis of minimization of the error function, generation of optimal weight values is carried out. Proposed R-JENN model is used in this paper to control the reactive power in the wind power system. R-JENN is well trained so as to constrict the violations in voltages and maintain the reactive power limit by means of proper modification in their control variables. STEP

FUNCTION

DESCRIPTION

1.

Define the JENN

2.

Train the JENN

3.

Test the JENN

4.

Optimize the JENN

5.

Test the R - JENN

Build the JENN: Determine number of nodes in each layer Initialize the input vector Set the necessary control parameters. Implement the desired training: Calculate the net input Calculate the outcome at output layer Calculate the error value Update the weights Conduct the learning until appropriate performance is achieved. Test the trained JENN with respect to meet the performance criterion stated in Eqn. (2). Build the R-JENN: Enable momentum factor variation between hidden and context layer. Enact the D-GABC rule. Input the trained values from JENN into D-GABC rule. Return the optimized weights of the JENN model and PID gain values. Compute the response of the system for the sixthorder transfer function given in Eqn. (1) with the designed new R-JENN controller model.

5 Numerical Simulation and Output Response of the DFIG Proposed D-GABC algorithm-based R-JENN model is simulated in MATLABSIMULINK environment, and the corresponding simulation results are presented in this section. Table 1 shows the design parameters of new R-JENN model employed in tuning the standard PID controller.

72 Table 1 Design parameters of the proposed R-JENN model

Table 2 Gains employing the proposed R-JENN model

S. N. Deepa and N. Rajasingam Parameters of R-JENN model No. of input neurons

3

No. of output neurons

3

No. of hidden layers

1

No. of hidden neurons

4

Activation function

Sigmoid

learning rate

0.2

Number of iterations

100

Number of trials taken

36

Momentum factor (based on trials)

0.7

Controllers

Proportional gain (KP )

Integral gain (KI )

Derivative gain (KD )

Tuned PID controller

12.3675

7.0821

0.1286

Proposed R-JENN controller

26.5964

4.0054

0.0165

Basically, PID controller is tuned conventionally employing Ziegler—Nichols tuning approach. Appropriate gain values attain in reduction of settling time, rise time, and steady-state error of the considered system model. Table 2 shows the obtained gain values by employing the proposed R-JENN-PID controller model for the related DFIG system. The performance criterion as stated in Eq. (2) was evaluated during the generation of D-GABC, and its value over the number of generations is denoted in Table 3. It is observed from Table 3 that over the generations, the performance Table 3 Performance criterions evaluated with the proposed R-JENN model

Number of generations

Achieved performance criterion

10

4.0007

20

3.8692

30

2.7750

40

2.4926

50

2.2286

60

1.6954

70

1.6126

80

0.7691

90

0.3547

100

0.1724

Regulated Jordan—Elman Neural Network-Based Controller Model …

73

criterion value (J I ) has got minimized. At the end of 100 generations, this fitness value is noted to be ‘0.1724’. Based on the gain values computed employing the proposed R-JENN controller model, the step-response of the DFIG system is as shown in Fig. 4, and Table 4 describes the comparison of the step-response characteristics for the DFIG on employing the designed controllers. From Table 4, it is noted that proposed R-JENN considerably reduces the steady-state values than that of the other controller models. The best performance criterion value (J I ) obtained by employing Eq. (2) over a certain number of trials for proposed R-JENN controller model along with considered controller design approaches is given in Table 5 and the proposed R-JENN model significantly minimized the performance criterion than that of other methods. If the PID controller is tuned accurately, it achieves better results. Due to this reason only, widely for applicability of controllers, PID controller is suggested and practically used successfully. Hence, in this work, neural controllers are designed,

Fig. 4 Response of the system using proposed R-JENN model

Table 4 Comparison of the step-response characteristics for the R-JENN model Characteristics methods

Settling time

Rise time

Peak overshoot

Peak time

Conventional PID

0.5291

0.0152

15.1985

0.1419

Proposed R-JENN

0.3301

0.0084

0

0.0581

74 Table 5 Comparison of the performance criterion values for the R-JENN model

S. N. Deepa and N. Rajasingam Controller Methodologies

Best performance criterion value

ABC approach [16]

0.4956

GABC approach [17]

0.2759

D-GABC approach [15]

0.2029

Proposed D-GABC-RJENN approach

0.1724

and effective control action is initiated on the system model to attain better results. Conducted performance analysis shows that the proposed D-GABC-based JENN as employed for controller modelling of DFIG-based grid-connected wind systems proves their effectiveness over the conventional controller models [18, 19]. The applicability of density-based GABC algorithm into the JENN model for weight tuning has resulted in stabilizing the regulated weight values so as to prepare the neural network converge at a faster rate.

6 Conclusion This paper dealt with D-GABC algorithm-based multilayer perceptron model, namely regulated Jordan—Elman neural network (R-JENN) for the purpose of evolution of DFIG controller model in wind energy generation. Developed neural controller model was applied with conventional PID controller in order to effectively drive the DFIG model. By employing MATLAB environment, devised multilayer perceptronbased controller model was simulated, and the effectiveness of proposed controller structure was validated based on the attained simulated output responses. It should be observed from the simulation plots that the developed R-JENN controller model has minimized the performance criterion to a substantial extent. It proves the performance of the proposed R-JENN controller model and paves its way for effective controller implementation for DFIG-based grid-connected wind energy systems.

References 1. Kalogirou SA (2000) Applications of artificial neural-networks for energy systems. Appl Energy 67(1–2):17–35 2. Li G, Shi J (2010) On comparing three artificial neural networks for wind speed forecasting. Appl Energy 7(87):2313–2320 3. Sivanandam SN, Deepa SN (2007) Principles of soft computing, 2nd edn. Wiley, India 4. Jordan MI (1986) Serial order: A parallel distributed processing approach. Tech. Rep. No. 8604. University of California, Institute for Cognitive Science, San Diego 5. Elman JL (1990) Finding structure in time. Cognitive science 14(2):179–211

Regulated Jordan—Elman Neural Network-Based Controller Model …

75

6. Kaloi GS, Wang J, Baloch MH (2016) Active and reactive power control of the doubly fed induction generator based on wind energy conversion system. Energy Reports 2:194–200 7. Ebrahimkhani S (2016) Robust fractional order sliding mode control of doubly-fed induction generator (DFIG)-based wind turbines. ISA Trans 63:343–354 8. Yang B, Zhang X, Yu T, Shu H, Fang Z (2017) Grouped grey wolf optimizer for maximum power point tracking of doubly-fed induction generator based wind turbine. Energy Convers Manage 133:427–443 9. Song G, Zhang C, Wang X, Huang W (2017) Simplified method of doubly fed induction generator. J Eng 2017(13):1200–1205 10. Abouheaf M, Gueaieb W, Sharaf A (2018) Model-free adaptive learning control scheme for wind turbines with doubly fed induction generators. IET Renew Power Gener 12(14):1675– 1686 11. Xiong L, Li P, Wu F, Ma M, Khan MW, Wang J (2019) A coordinated high-order sliding mode control of DFIG wind turbine for power optimization and grid synchronization. Int J Electr Power Energy Syst 105:679–689 12. Justo JJ, Mwasilu F, Jung JW (2015) Doubly-fed induction generator based wind turbines: a comprehensive review of fault ride-through strategies. Renew Sustain Energy Rev 45:447–467 13. Krause PC, Wasynczuk O, Sudhoff SD, Pekarek S (2002) Analysis of electric machinery and drive systems, vol 2. IEEE Press, New York 14. Ko HS, Yoon GG, Kyung NH, Hong WP (2008) Modeling and control of DFIG-based variablespeed wind-turbine. Electric Power Syst Res 78(11):1841–1849 15. Rajasingam N, Rasi D, Deepa SN (2019) Optimized deep learning neural network model for doubly fed induction generator in wind energy conversion systems. Soft Comput 23(18):8453– 8470 16. Karaboga D (2005) An idea based on honey bee swarm for numerical optimization. Technical report-tr06, Erciyes University, Engineering Faculty, Computer Engineering Department 200, 1–10 17. Xiang WL, Li YZ, Meng XL, Zhang CM, An MQ (2017) A grey artificial bee colony algorithm. Appl Soft Comput 60:1–17 18. Abdulkarim SA (2016) Time series prediction with simple recurrent neural networks. Bayero J Pure Appl Sci 9(1):19–24 19. More A, Deo MC (2003) Forecasting wind with neural networks. Marine Structures 16(1):35– 49

BESS-Based Microgrid with Enhanced Power Control and Storage Management Shanmugam Kalpana and Gogineni Pradeep

Abstract In order to meet the continuous growing electricity demand with reduced environmental pollution, renewable energy resources are in extensive use, and the traditional energy generation and distribution is being replaced by the microgrid. An increasing share of renewables penetrations and changes in power/energy consumption patterns in the residential/industrial sector are impacting the grid code tightening requirements for the reliability and stability of the power system. BESS can be viewed as a valuable and reliable asset that can provide more added benefits, flexibility, and stability to the microgrid system in both islanded and grid-connected operation conditions. In this paper, the microgrid with BESS is analyzed in different operating conditions. The intermittent behavior of renewable energy can result in number of operational challenges including frequency and voltage fluctuations in microgrid. BESS is used to counteract the intermittent nature of renewables, thus by providing reliable, stable power. This paper proposes a distributed secondary control for islanded and grid-connected microgrids with BESS. Droop control is used to perform autonomous power management among sources in microgrids (MGs). This approach has the potential to improve system reliability, stability, and ease of management. The proposed secondary control requires only simplified communication protocol and a dedicated communication network. The control algorithm implemented in microgrid system is tested in different scenarios by means of software-in-the-loop, hardware-in-the-loop, etc. However, in this paper, we present performance of the system through software-in-the-loop test results for the islanded, grid-connected, and transitions operating scenarios. Keywords Energy storage · Grid connected and islanded · Microgrids resilience

S. Kalpana (B) · G. Pradeep ABB GISL Pvt Ltd, Bengaluru, KA 560048, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_7

77

78

S. Kalpana and G. Pradeep

1 Introduction A microgrid is a small-scale electric grid that combines power from distributed energy resources (DER) such as combined heat and power plants (CHP), dieseland gas-powered gensets, renewable sources (PV, wind), energy storages (battery and flywheel), hydroelectric, and tidal. MG should have the capability to control and coordinate all the connected generation assets to meet the requirements of load (industrial, consumer, residential). Due to the transformation toward de-carbonization, we need to primarily rely on MG, which are autonomous energy supply systems that are efficient, reliable, and environmentally friendly. The MG systems makes social and economic development possible, changing lives by providing access to affordable, reliable electricity. The introduction of power can be life changing for rural areas---a catalyst for social and economic development of the country. Basic living standards are improved with the introduction of water pumps, refrigeration for perishables and medicines, light to see by at night, and a reduced reliance on expensive. In other way, frequent power outages disrupt both commercial and industrial activities, resulting in economic loss and factory production in-effective and prevent business success which is potentially dangerous and challenge to the country economy. So, MG is the one which can provide electricity to remote villages/islands which are not connected to the main grid, or reliable backup power for small commercial and industrial facilities connected to unreliable grid supply. MG should also be capable to operate on on-grid (grid connected) or off-grid (islanded). Also, from grid codes perspective of different countries, MG control functions should make it capable to seamlessly connect and disconnect from main grid during the potential grid fault or emergency, and this is increasingly important feature. Integrating the renewable energy to MG and control, coordinating of the fluctuations and high penetration of renewables is another increasingly important feature. This supports uninterrupted power as well as encourage clean energy use. Digital automation and control systems intelligently coordinate DER and loads for the MG to function efficiently.

2 Microgrid System and Controls 2.1 Microgrid Architecture The centralized secondary control consists of one central secondary controller which collects the information from all the primary controllers in the systems and sends back the calculated set points to the primary controllers using dedicated communication channels. The centralized control approach requires point-to-point communication, which adds complexity to the system and compromises its reliability due to a single point of failure issue [1]. Alternatively, distributed secondary control method can be used. Due to its attractive features, they have recently drawn a lot of attention in MG research community [2]. In the distributed control architecture, secondary control

BESS-Based Microgrid with Enhanced Power Control and Storage …

79

Fig. 1 MG system overview (all generation assets connected)

consists of equivalent number of primary controllers. Each secondary controllers gets information through the dedicated communication network from all the other secondary controllers, and the set points are evaluated smartly based upon the gathered information. Different energy storage technologies are available [3, 4], and among them BESS is dealed in this paper (Fig. 1).

2.2 Coordinated Primary and Secondary Control In islanded microgrids, the output power from BESS and other units should be controlled in a coordinated way concerning batteries state of charge (SoC) condition so that to keep balance between power generation and consumption and at the same time prevent BESS from over charge scenario. This coordinated objective is fulfilled based on two-level hierarchical control structure: primary level with regards to power electronics control and secondary level for bus variable regulation. All the secondary controllers need to be mutually synchronized and should have dedicated communication protocol. Primary control is responsible for individual converter power, voltage, and frequency regulation. Secondary controllers take the responsibility to perform power quality regulations (harmonics) to manage voltage/frequency deviations, unbalances, etc. Also, it provides support for synchronizing between the MG and external grid. Tertiary control deals with the advanced functions related to efficiency and economic enhancements which constitute a higher management level, mostly referred as the energy management system (EMS). All the secondary controllers are connected to each other by Peer to Peer network in real world, by a table called (MG table) in the simulation world. Primary control response time will

80

S. Kalpana and G. Pradeep

Fig. 2 MG system control layers architecture

be in microseconds, and secondary will be in milliseconds. The primary coordinated control aims at power regulation of all units based on different SoC scenarios of BESS (Fig. 2).

2.3 Droop-Based Secondary Control Many control strategies, such as droop control and master–slave control have been implemented worldwide to operate parallel-connected inverters for load sharing in microgrid. Among these methods, the droop control technique has been widely accepted in the scientific community due to the absence of critical communication links among parallel-connected inverters to coordinate the DER units within a microgrid. The droop control regulates frequency (f ) and voltage (V ) which causes the active (P) and reactive power (Q) sharing. In the inductive transmission line, active power (P) mainly depends on the power angle (δ). Reactive power (Q) depends on the voltage difference (v). So, from control perspective, power angle (δ) is used to control P, and voltage difference (v) can be used to control Q. In the microgrid, the droop control strategy uses the droop characteristics of traditional power system, by changing the output of active and reactive power to control the frequency and amplitude of the output voltage, so that microgrid system can work on stabilizing voltage and frequency point in island operation mode (Fig. 3).

2.4 Battery Energy Storage System (BESS) BESS provides increased flexibility of microgrid system by storing the excess energy during off peak demand and delivering them during on peak period. Energy storage

BESS-Based Microgrid with Enhanced Power Control and Storage …

81

Vmax f0 fN

VN Vmin

fmin

PN

Pmax

(a) f-P droop

Qmin

Qmax

(b) V-Q droop

Fig. 3 Drooping curve: a F–P droop b V –Q droop

technology can vary the real power according to the changing demand. BESS also supports to increase the system efficiency and makes the system more economic [5]. BESS can provide power system stability, and it also can improve the power quality. BESS has an inverter which can transform the DC voltage from the battery to the AC voltage needed for the grid or microgrid and vice versa. The inverter has a voltage source converter and a pulse width modulator [6, 7]. It can operate in both virtual generator mode (VGM) and grid supporting mode (GSM). Two current parameters, d-axis and q-axis, can be controlled to manage the real and reactive power supplied by BESS, respectively. Placement of energy storage device has significant impact on frequency recovery performance of microgrid [8] (Fig. 4).

Fig. 4 Annual cumulative installed capacity of utility-scale BESS (i.e., > 10 MW) and their growth projected based on the projects announced until August 25, 2015 [6]

82

S. Kalpana and G. Pradeep

Fig. 5 Operating area in the PQ curve according to the PQ priority

2.5 BESS Power Capabilities BESS is suitable for numerous applications due to its unique characteristics such as capability of working in both the generation and load modes, fast and precise response to the control signal (V, F), capability of providing reactive power services (both supply and consumption), and relatively high efficiency [9]. The impact of reactive power on the active power availability depends essentially on the PQ curve of BESS [1]. The impact of active power on the availability of reactive power depends not only on the PQ shape but also on the frequency quality and the operating parameters in the frequency regulation task (such as activation time) [10]. The right priority of tasks that delivers the maximum profit to a BESS operator is sensitive to the system market prices and rules, as well as the needs of the local distribution grid [3]. More analyses and intelligence need to be added to carry out also to define the optimum priority of tasks (Fig. 5).

3 Proposed System Design 3.1 Proposed System and Simulation Modelling Simulations have been performed in DigSILENT power factory. DigSILENT power factory software used for power system modelling and simulation. It has the functionality to conduct steady-state, time-domain (quasi-dynamic, RMS, and EMT) RMS modelling. These simulations are commonly used throughout the world for most of

BESS-Based Microgrid with Enhanced Power Control and Storage …

83

the grid analysis/study. These simulations are also referred to as (dynamic) stability simulations, and the advantage of using RMS simulations is that the simulation speed can be increased/improved significantly. We have used DIgSILENT simulation language (DSL) for model definition and composite frames for defining the MG resources. DIgSILENT also provides the flexibility to either simulate with a fixed time step or a variable time step (Fig. 6). The primary and secondary controllers of all MG sources are modelled as composite frames in the DIgSILENT. The secondary control is linked by using dll. At any time, the user can switch ON/OFF the assets by switching off their corresponding frame defined.

Fig. 6 Simulation model of the MG system

84

S. Kalpana and G. Pradeep

3.2 Operational Control of Proposed System Distributed control architecture with dedicated peer-to-peer communication between the secondary controllers is proposed which provides more flexibility, simplicity, scalability, and modularity to the microgrid [11–13]. Control logic can be developed and dispersed across the secondary controllers at any time. This allows enhancements that must be made to meet new market requirements and/or the addition of new resources or controllers or field devices, to be made efficiently and accurately, easing and simplifying updates, and new equipment integration in to the existing MG system. The set points are calculated in the secondary controllers based upon the primary controllers’ actual values. The actual microgrid system should follow the single order response for attaining the required set point. The secondary controller continuously calculates the set point such that the primary controller response should be in smoother way. All the secondary controllers are communicating with each other through the dedicated peer-to-peer communication network in real world, through the MG table in the simulation environment.

4 Design Specifications

Table 1 Ratings of the microgrid assets

Microgrid asset

Power system rating

Diesel generator 1

1250 kVA

Diesel generator 2

1250 kVA

Photovoltaic

1250 kVA plant

BESS

1000 kW/500 kWh (Power converter/battery)

Feeder load

Feeder-3000 kVA

Network

5 Control Strategy and Validation The microgrid consists of two diesel generators, one PV, one BESS, one feeder load, and grid connection. So, total six secondary controllers are present in the system. All the primary controllers’ information is received through the dedicated communication network. Microgrid stabilization, power limiting, renewable smoothing, smooth transition, and peak shaving features are explained for different operating conditions

BESS-Based Microgrid with Enhanced Power Control and Storage …

85

of the system. The BESS is used to stabilize renewable energy, and it is discussed in [18]. The interaction of BESS for renewable is discussed in [19].

5.1 Voltage and Frequency Stabilization (In Islanded) BESS secondary control is having the voltage and frequency support functionality in the islanded network. Whenever the load got decreased, the territory controller switches off all the diesel generators in the system. The BESS acts as grid-forming device momentarily, and voltage and frequency reference values are provided to the system. BESS supports the frequency and voltage while the diesel generators also present in the system. The frequency fluctuations are reduced due to the BESS controller support. BESS inertia is added to the diesel generators inertia, so that the system is stable for load variations [7].

5.2 Renewable Power Limitation (Islanded) When adding renewables sources to a MG, sometimes the control must limit the renewables power. This can be due to insufficient load to take all the power, insufficient storage power rating to accept all the excess power (storage may already be full SoC high), generator loading may go below the minimum, etc. This functionality is more important to make the system stable for the excess renewable power generation. BESS plays a vital role in integrating the renewables in MG system [14] and [15].

5.3 Renewable Power Fluctuations and BESS Requirement (In Islanded) The output from renewable is always a varying one without any warnings. This impacts the diesel generators operation and their efficiency. The energy stored in BESS can be used to counteract the variations in renewable energy sources. The renewable source and BESS could work in concert to create a total power output that is less variable. Although the BESS could be used to fill in all the valleys and cutoff all the peaks, creating a dead flat composite output, the microgrid can usually sustain power injection with some variation.

86

S. Kalpana and G. Pradeep

5.4 Peak Lopping This is the ability of the microgrid to ensure that the power flow to and from the external grid at the PCC is limited to certain threshold limit defined by the TSO. Nothing but controlling the aggregated load and generation power profile typically shaving and lopping are intended to be power functions delivered in a short amount of time while shaping and shifting are intended to the energy functions delivered for a long time. Peak lopping is described in [10].

5.5 Grid Connected to Islanded Transitions If microgrid can operate both on-grid and off-grid, it is necessary to manage the transition from on-grid to off-grid coordinating the microgrid assets, so that the grid breaker can be opened with the minimum power flow across it so to avoid unnecessary transients. Once the planned islanding procedure is triggered, the set points are calculated such that the real and reactive power flow at the PCC to be attained to zero values smoothly, so that diesel generators takes up the deficit power in ramped way. It is finally worth to highlight that because the planned islanding action has significant system level effects, it is often required to manage additional checks, timers, warning, and alarms in the application layer which might need additional input, output, and parameters.

6 Simulation Results and Analysis With respect to the design specifications and components sizing, a basic model inspired by the idea of pulse production is simulated in DigSILENT. The simulated model’s specifications are given in Table 1.

6.1 Voltage and Frequency Stabilization (in Islanded) The MG system is considered as islanded one. The real and reactive power load values are changed with step change. The frequency and voltage fluctuations are plotted. The BESS control along with generator controllers reduces the voltage and frequency drops for the real and reactive power load variations. The stop command is provided to the diesel generator controllers. The system is running with the BESS controller without any disturbances. BESS controller acts as voltage and frequency reference mode while the diesel. Generators are in off condition. The PV converter also runs smoothly in this scenario (Fig. 7).

BESS-Based Microgrid with Enhanced Power Control and Storage …

87

Fig. 7 Voltage and frequency stabilization in an islanded MG system

6.2 Renewable Power Limitation (Islanded) In the islanded microgrid system, the PV controller send the set points to the PV inverter based upon the load requirement, generator minimum loading, and BESS charging capability. Generators total min loading is 600 kW. BESS charging capability is at 800 kW. The active power load value is 1200 kW. Initially, the PV was limited to its maximum capability of 1000 kW. 400 kW power is used for BESS charging. The load value is reduced to 400 kW. The renewable power is limited to 600 kW value. Since the BESS is having maximum charging capability of 800 kW, further load was also reduced to the negative value for verifying the renewable power limiting. The PV controller is sending the set point correctly as per the operating condition (Fig. 8).

6.3 Renewable Power Fluctuations and BESS Requirement (In Islanded) [Smoothing] The Islanded microgrid system behavior for the renewable power fluctuations is tested. Initially the PV is supplying the power 600 kW. The PV power was changed suddenly to 200 kW and bring back to 600 kW. The Diesel generator and BESS

88

S. Kalpana and G. Pradeep

Fig. 8 Renewable power limitation feature on secondary control layer in an islanded MG system

active power variations are shown in Fig. 1. The deficit power at one step is supplied by the BESS and slowly ramped down to zero value. So that the deil generators do not take this sudden stress. The excess power at second step is absorbed by the BESS and slowly ramped up to zero value (Fig. 9).

6.4 Peak Lopping The grid-connected MG system is considered for validating the peak lopping behavior. The active power export and import limits considered at PCC are 200 kW and −200 kW. BESS charging capability is reduced to 200 kW assuming that battery is nearer to its full charge. Initially, the load power is small compared to the available PV and diesel generators active power. The microgrid is exporting the active power upto its limit of 200 kW. The PV power is reduced to 700 kW. The microgrid control system first reduces the grid export power, and later it reduces the BESS charging power for supporting the load power. At the second step, the grid and BESS charging power got reduced. At the third step, the PV power is further reduced. In this step, the grid starts importing active power and BESS charging power reduced to zero value. At the fourth step, the load was reduced in ramped way. Grid-importing power reduced slowly and started exporting the active power upto its limited value. BESS charging power also increased slowly. Microgrid secondary controller monitors the system

BESS-Based Microgrid with Enhanced Power Control and Storage …

89

Fig. 9 Renewable power fluctuations and BESS counteraction in an islanded MG system

operating conditions and sends the set points to all PV, diesel generator, and BESS controllers to maintain the grid active power within the allowable limits (Fig. 10).

6.5 Grid Connected to Islanded Transitions The grid-connected MG system is considered for validating the smooth transition from grid connection to the islanded one. Initially, the MG system is having the less PV power, and it is importing the power from the grid. At 320s, the grid transition signal is initiated from the customer. At this step, microgrid control system sends the command to BESS controller such that it takes up the power from the grid support in the ramped manner. Once the active power and reactive power reaches the zero values, the opening command is given to the grid circuit breaker. At 350s, the microgrid system is in islanded stage. BESS controller sends the power set point as zero, so that the required load power apart from the PV controller is supplied by the diesel generator. The microgrid control system makes the smooth transition in the system. If the grid is tripped suddenly, then also the instant power will be supplied from the BESS, and its final set point is attained in ramped manner (Fig. 11).

90

Fig. 10 Peak lopping in an islanded MG system

Fig. 11 Grid connected to islanded transition in an MG system

S. Kalpana and G. Pradeep

BESS-Based Microgrid with Enhanced Power Control and Storage …

91

7 Conclusion Over the last few years, BESS has been established as an essential component for renewable integrated microgrid. The analysis of power management and control strategies of islanded/grid-connected microgrid with storage is presented in this paper. This paper shows how BESS provided the flexibility and, V, stability to the microgrid using the distributed control architecture and also how it counteracts for fluctuations in renewables, so that the power will be reliable and stable. The simulation results have been discussed for seamless transition from islanded to grid connected. In this paper, we present performance of the system through software-inthe-loop test results. Based on the observed benefit of using BESS in microgrid, we can build more efficient, cost-effective microgrids. Also, ABB’s microgrid projects are included in the reference sections [4, 16, 17].

References 1. Shaila A, Tareq A (2017) Impact of battery energy storage system on post-fault frequency fluctuation in renewable integrated microgrid. In: International conference on electrical, computer and communication engineering. ECCE, Cox’s Bazar, Bangladesh, pp 594–598 2. Yang R-H, Huang W (2015) Micro-grid structure and operational control. Int J Control Auto 6:65–76 3. Srinivas BK, David X (2016) Optimal capacity and placement of battery energy storage systems for integrating renewable energy sources in distribution system. In: National Power Systems Conference NPSC. Bhubaneswar, India 4. ABB Reference projects, https://new.abb.com/news/detail/2290/the-energy-example-of-rob ben-island-mandelas-prison 5. Qin L-J, Yang W-T (2017) Micro-grid droop control strategy and simulation. In: Advances in electrical and computer engineering Beijing, China, pp 74–80 6. Wang G, Konstantinou G, Townsend CD, Pou J, Vazquez S, Demetriades GD, Agelidis VG (2016) A review of power electronics for grid connection of utility–scale battery energy storage systems. IEEE Trans Sustain Energy 7(4):1778–1790 7. Nadezda B, Pirjo H, Atte P, Juha K, Kristiina S (2017) Multi-objective role of a bess in an energy system. In: CIRED Workshop, computer and communication engineering. ECCE, Ljubljana, Bangladesh. 8. Chawengsak W, Paramet W (2018) Control strategy for seamless transition of microgrid using battery energy storage system. In: 53rd international universities power engineering conference. UPEC, Glasgow, Scotland 9. Nicholas M, Devon M, Jim R, Paul M, Erik K (2010) Utility scale battery energy storage systems. In: IEEE PES general meeting 2010. Rhode Island, USA 10. Prashant KP, Atma RG (2018) Battery energy storage system. In: 4th international conference on computational intelligence & communication technology. ICICT, Ghaziabad, India, pp. 594– 598 11. Zhang J, Wang Q, Hu C, Rui T (2017) A new control strategy of seamless transfer between gridconnected and islanding operation for micro-grid. In: IEEE conference on industrial electronics and applications. ICIEA Siem Reap, Cambodia, pp 1729–1732 12. Katiraei F, Iravani MR (2006) Power management strategies for a microgrid with multiple distributed generation units. IEEE Tran Power Syst 21(4):1821–1831

92

S. Kalpana and G. Pradeep

13. Tomislav D, Wu D, Qobad S, Lexuan M (2017) Distributed and decentralized control architectures for converter-interfaced microgrids. Chinese J Electri Eng 3(2):41–52 14. Abdullah A.A, Hussein MK, Al-Masri ME (2019) Integration of renewable energy sources by load shifting and utilizing value storage. IEEE Trans Smart Grid 10(5):4974–4984 15. Noriko KYI (2012) Overview of battery energy storage systems for stabilization of renewable energy in Japan. In: International conference on renewable energy research and applications, 2012, ICRERA. Nagasaki, Japan, pp 594–598 16. ABB Reference projects, https://www.abb.com/microgrids 17. ABB Reference projects, https://new.abb.com/news/detail/13052/abb-state-of-the-art-microg rid-to-power-african-hub-of-the-international-committee-of-the-red-cross

Adaptive Reserve Estimation Technique for PV Systems Under Dynamic Operating Conditions Pankaj Verma , Tarlochan Kaur , and Raminder Kaur

Abstract As the PV systems are static in nature and lack inertial properties, the situation demands for the ways to couple the system frequency with the active power, because increasing penetration of the static PV devices has caused a serious threat to the stability of the power system. Many methods have already been proposed to enable the inertia capabilities in PV systems, one of them is to keep a reserve in the system by deloading. In deloading, the reserve is kept as a fixed or a constant value, which might lead to challenges in the operation of the system under dynamic conditions. In this paper, an adaptive reserve estimation technique is proposed in which the power reserve is flexible and is determined using a fuzzy logic controller. The fuzzy rule base is developed by considering the irradiance and ambient temperature of the PV array. The control aims at making the reserve technique flexible to dynamic conditions. The control has been modeled in the Simulink and the working is tested by simulating the system for a dc load. Keywords MPP (maximum power point) · RPT (reference power tracking) · STC (standard test conditions)

1 Introduction Due to the high carbon emissions from the conversions of conventional energy sources, the power sector has experienced a spike in the installation of the renewable energy sources since a decade. The solar photovoltaic (PV) systems which are believed to be the most favorite amongst the category have also shown a similar growth pattern. However, the static nature of the PV systems has caused many challenges to the system stability in the grid connected configuration. As these systems lack inertia, under high PV penetrations, the small system disturbances may result in a complete loss of synchronism. As evident from [1–5], upon a system disturbance,

P. Verma (B) · T. Kaur · R. Kaur Punjab Engineering College (Deemed To Be University), Chandigarh 160012, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_8

93

94

P. Verma et al.

a high PV penetration causes a high deviation of the generator speed, lower system damping characteristics, higher frequency deviation, lower system stability, etc. To cope up with this problem, one easy way is to use the storage batteries which may release the available power into the grid for regulating the frequency. However, batteries have a high purchasing, operation and maintenance costs and an extra hardware is required. Another way is to generate an active power reserve in PV’s which could be utilized during an event of a sudden load change in the system. There are many approaches used by the authors to generate the power reserve, like in [6], the hill climbing algorithm is modified to follow the commanded power. In [7], Newton quadratic interpolation algorithm is modified to trace the fixed reference level. In Ref. [8], a PI-based deloading controller is used for operating the PV system above maximum power point voltages, resulting in a system reserve. A neural network bases mechanism is used in [9] for generating a power reference corresponding to the 10% reserve. A fuzzy logic controller for computing the reserve based upon the system inertia and PV penetration levels is discussed in [10]. A simple incremental conductance algorithm is modified in [11] to trace the reference power level. A modified open circuit voltage method for creating a reserve in the system is suggested in [12]; here, the range of gain ‘k’ is changed from 0.71–0.78 to 0.8–0.95. Similarly, a new limited power point tracking method, based on the perturb and observe approach, is proposed in [13]. Some of the methods [6, 7, 11–13] have kept the reserve as fixed value, while [8] and [9] have considered the reserve to be a fixed percentage of maximum available power and [10] have obtained the reserve using a fuzzy controller. For the methods keeping the power reserve as a fixed value, the reserve remains fixed even if the available PV power is abundant or scarce. A lower availability of PV power may lead to an operation failure for the case of fixed reserve conditions. Hence, the percentage reserve is more appropriate. However, these methods may become problematic under the situation illustrated in Fig. 1. In Fig. 1, as the PV power drops to a significant lower value, there is a need to adjust the percentage or the fractional reserve in the system. If the system is operated on the same reserve for curve ‘b’, the connected loads may get compromised. To filter out the limitations of these methods, as highlighted above, adaptive reserve estimation techniques is developed in this paper. Under this technique, a Fig. 1 PV curves under different operating conditions

Adaptive Reserve Estimation Technique for PV Systems Under …

95

fuzzy logic controller is used to compute the reserve in the system having the irradiance and ambient temperature as inputs. The reference power is traced using the proposed reference power tracking algorithm (RPT). The working of the proposed controller is verified by simulations in the MATLAB software. The paper is further divided into various sections, Sect. 2 discusses the proposed adaptive reserve control strategy, the simulations results have been presented in Sect. 3, the discussions and comparison of the proposed techniques with the report research are carried out in Sect. 4, and the conclusions are given in Sect. 5.

2 Adaptive Reserve Estimation Technique The multistage control for the PV system is shown in Fig. 2. In the first stage, the output from the PV generator is connected to a dc/dc converter for tracing the reference power; here, the converter is not operated in MPPT mode, and the duty cycle is computed using the proposed reserve control strategy. In the second stage, a multiple feedback loop control strategy [14] is used for the inverter control. The inverter serves the purpose of dc to ac conversion and also maintains a constant voltage across the feeder .

Fig. 2 Multistage control of PV system with proposed reserve controller

96

P. Verma et al.

Table 1 Fuzzy system rules TdG

VH

H

M

L

VL

VL

VL

VL

VL

L

L

L

VL

VL

L

L

L

M

L

L

M

M

M

H

L

M

M

H

H

VH

M

M

H

VH

VH

2.1 Reserve Controller In the reserve controller, the fractional active power reserve ‘r’ is obtained using a fuzzy controller. The inputs to the controller are irradiance G and temperature deviation T d (deviation from the STC condition, 25 °C), since the changes in the irradiance and the ambient temperature are not fixed and occurs over a nonlinear scale; hence, the fuzzy logic control is preferred here as it handles the nonlinearity’s most efficiently. After the process of fuzzification, the inputs are divided into five classes which are very low (VL), low (l), medium (M), high (H), and very high (VH). The detailed distribution of the membership functions can be followed from the appendix. The fractional reserve ‘r’ is the output obtained after the defuzzification process; the reserve of the system varies from 0 to 0.2. The maximum allowed reserve for the developed control is 20%. The fuzzy rules linking the two inputs with the output are summarized in Table 1, the rules are developed by using the fact that when the operating parameters of the PV array deviate from the standard test conditions (STC), the system reserve needs to be decreased so as to compensate the connected load. The surface view of the rules is in Fig. 3. Once the fractional reserve command is obtained, the reference power Pref is obtained using Eq. (1) Pref = Pmax (1 − r )

(1)

where Pmax is the maximum available power. The obtained reference power is followed with the help of a suggested reference power tracking algorithm which is shown in Fig. 4. In the method, during starting, RPT algorithm traces the maximum power, after few seconds the commanded reference power is traced. If the commanded power is above maximum available power, then the maximum power is traced. If the reference power is below the MPPT, the PV array voltages are increased to reach the required power. If in this process the PV power overshoots the required power, the PV array voltages are then decreased to trace the required operating point. Hence, the PV system is deloaded by operating the PV on the right side of the MPPT on the power vs. voltage (P–V) curve. Most of the available methods in the field of deloading have preferred the right-hand side operation, while few have favored the left-side operation. The left-side operation gives rise to many complications in the system.

Adaptive Reserve Estimation Technique for PV Systems Under …

97

0.15

r

0.1

0.05

20 15

1000 800

10

Tdev (deg cel.)

600 400

5 0

200 0

G (W/m2)

Fig. 3 Surface view for fuzzy rules

Fig. 4 Proposed reference power tracking algorithm

3 Simulation Results To verify the effective working of the reserve control scheme, the simulations have been carried out on the DC system with a dc load of 2 kW, the detailed system parameters are given in Table 3 in appendix. The simulations are carried out in the

98

P. Verma et al.

dynamic operating conditions, initially in the first scenario, the irradiance is varied, and in the second scenario, both irradiance and ambient temperature are altered. The simulation results for the first scenario are presented in Fig. 5. Initially from t = 0 to t = 0.5s, the MPP is followed Fig. 5a, during this period the operating conditions are maintained at STC. As the irradiance falls from 1000 to 800 W/m2 and then to 600 W/m2 as illustrated in Fig. 5e, the system starts following the reference power levels of Fig. 5b. The obtained system reserve is shown in Fig. 5d; the reserve adjusts as according to the created rule base of the fuzzy controller. The PV output power with variable reserve (obtained from proposed controller) and a fixed reserve of 0.2 is compared in Fig. 5a, clearly the system is more flexible with the proposed reserve control. Similarly, the array voltages for the two reserve conditions are shown in Fig. 5c, the system power is reduced by increasing the voltages above MPP, hence justifying the right-hand side operation. The results obtained by varying both, the irradiance G and temperature T, are presented in Fig. 6; here, the system is simulated for the same irradiance variations as in Fig. 5e, while the temperature variations are shown in Fig. 6c. Figure 6a shows the comparison of the PV power for the variable reserve and the fixed reserve; it can be noticed here that the power gap between the two PV power (for fixed reserve and for variable reserve) has increased in Fig. 6a as compared to Fig. 5a. The system power reserve is shown in Fig. 6b, for this case the reserve falls to a value of 0.06, while for the previous case the minimum reserve was around 0.10

4 Discussions The suggested adaptive reserve estimation control for PV systems predicts the fractional reserve by considering the operating parameters of the PV array, instead of operating at a fixed reserve level. A comparison of the proposed technique with some of the popular reserve methods is presented in Table 2. As discussed earlier, the methods [6, 7, 11–13] which are keeping the reserve as a constant or a fixed value are more prone to conflicts associated with operation and control of the system. Techniques having the reserve as a fixed percentage of maximum power [8, 9] are more preferred, but these techniques have some limitations as discussed in introduction section. In [10], the reserve is determined by considering the average inertia and the level of PV penetrations, while the operating parameters for the PVs were not considered. Hence, in above technique, the reserve control may ask for a value of reserve which may cause a serious threat to the stability of the connected system. The amount of reserve provided by various control schemes is given in column 5 of Table 2. The system stability is affected by the flexibility of the provided reserve, as in Ref.[7], the reserve is kept as 0.03 MW for all the operating conditions of PV, here a situation may prevail where the power from PV reduces drastically, and the operation on the same reserve will result in higher frequency oscillations. Hence, in the proposed control, reserve is changed with operating conditions.

Adaptive Reserve Estimation Technique for PV Systems Under …

99

Fig. 5 Simulation of the PV system with proposed reserve control under variable irradiance, a PV output power, b Reference power, c PV array voltages, d Fractional reserve, and e Irradiance level

100

P. Verma et al.

Fig. 5 (continued)

5 Conclusions The reserves in the PV systems are generated so as to enable the inertial capabilities; however, the operation with a fixed reserve causes many operation challenges. In this paper, an adaptive reserve estimation technique for PV systems under dynamic operating conditions has been proposed and its effective operation has been verified. The verification of the proposed fuzzy control has been performed using simulations on the DC load. The system is simulated with a fixed reserve, and in the presence of proposed control, the power gap between the PV outputs highlighted the flexibility in the control. Overall, the developed control makes the PV systems with reserve more adaptive to the dynamic operating conditions.

Adaptive Reserve Estimation Technique for PV Systems Under …

101

Fig. 6 Simulation of PV system under variable irradiance and temperature, a PV power, b Fractional reserve, and c Operating temperature

102

P. Verma et al.

Table 2 Comparison of the proposed technique with reported methods S.No. Technique

Amount of reserve (r)

Generation of reserve using

Reserve flexibility (dictating system stability)

1

Altered hill climbing algorithm [6]

Constant (constant Commanded power)

Altered hill climbing algorithm

r = 50 W, for all operating conditions of PV

2

Altered Newton Fixed (constant quadratic reference power) interpolation (NQI) algorithm [7]

Altered NQI algorithm

r = 0.03 MW, for all operating conditions of PV

3

PI-based deloading controller [8]

Flexible (18% reserve)

Suggested PI controller

r = (0.18*MPP), for all operating cond of PV

4

ANN-based predictive reserve technique [9]

Flexible (10% reserve)

ANN and PI controller

r = (0.10*MPP), for all operating cond of PV

5

Power curtailment technique for PV’s [10]

Flexible (fuzzy controller output)

Suggested PI control

r α (system inertia), for all operating cond. of PV

6

MPPT ‘off’ mode algorithm [11]

Constant (const. reference power)

Altered INC algorithm

r = 0.4 p.u., for all operating conditions of PV

8

Virtual MPP traction [12]

Constant (reserve Altered open circuit r = 400 W, for all change with value of voltage method operating conditions ‘k’) of PV

9

Limited power point tracking (LPPT) algorithm [13]

Constant (const. reference power)

Suggested LPPT algorithm

r = 20 KW, for all operating conditions of PV

10

Proposed adaptive reserve control

Flexible (fuzzy control output)

Reference power traction algorithm

r = dependent on G and T d with max = (0.2*MPP)

Appendix See Fig. 7. See Table 3 .

Adaptive Reserve Estimation Technique for PV Systems Under …

103

Fig. 7 (a) Distribution of input ‘1 of fuzzy control. (b) Distribution of input ‘2 of fuzzy control. (c). Distribution of output of fuzzy control

104 Table 3 Simulation model parameters

P. Verma et al.

VOC

363 V

ISC

31.4 A

Pmax

8.5 KW

dc/dc source capacitance

100 µF

dc/dc inductance

5 mH

dc/dc switching frequency

3 kHz

Inverter link capacitance

1000 µF

Inverter switching frequency

20 kHz

LC filter capacitance

1.7 µF

LC filter inductance

20 mH

DC load

2 kW

References 1. Eftekharnejad S, Vittal V, Heydt GT, Keel B, Loehr J (2013) Impact of increased penetration of photovoltaic generation on power systems. IEEE Trans Power Syst 28:893–901 2. Pethe AS, Vittal V, Heydt GT (2014) Evaluation and mitigation of power system oscillations arising from high solar penetration with low conventional generation. In: Proceedings of North American power symposium (NAPS), pp 1–6 3. Hoballah A (2015) Power system dynamic behavior with large scale solar energy integration. In: Proceedings of 4th international conference on electric power and energy conversion systems (EPECS), pp 1–6 4. Bueno PG, Hernandez JC, Rodriguez JR (2016) Stability assessment for transmission systems with large utility-scale photovoltaic units. Renew Power Gener 10(5):584–597 5. Remon D, Cantarellas AM, Mauricio JM, Rodriguez P (2017) Power system stability analysis under increasing penetration of photovoltaic power plants with synchronous power controllers. IET Renew Power Gener 11(6):733–741 6. Watson LD, Kimball JW (2011) Frequency regulation of a microgrid using solar power. In: Twenty-sixth annual IEEE applied power electronics conference and exposition. (APEC), pp 321–326 7. Xin H, Liu Y, Wang Z, Gan D, Yang T (2013) A new frequency regulation strategy for photovoltaic systems without energy storage. IEEE Trans on Sust Energy 4(4):985–993 8. Zarina PP, Mishra S, Sekhar PC (2014) Exploring frequency control capability of a PV system in a hybrid PV-rotating machine-without storage system. Int J Elect Power Energy Sys 60:258–267 9. Sekhar PC, Mishra S (2016) Storage free smart energy management for frequency control in a diesel-PV-fuel cell-based hybrid ac microgrid. IEEE Trans. Neural Netw Learn Sys 27(8):1657– 1671 10. Rajan R, Fernandez FM (2019) Power control strategy of photovoltaic plants for frequency regulation in a hybrid power system. Elect Power Energy Sys. 110:171–183 11. Miloˇsevi´c M, Rosa P, Portmann M, Andersson G (2017) Generation control with modified maximum power point tracking in small isolated power network with photovoltaic source. In: IEEE Power Engineering Society General Meeting, pp 1–8 12. Pappu VAK, Chowdhury B, Bhatt R (2010) Implementing frequency regulation capability in a solar photovoltaic power plant. In: North American power symposium, pp 1–6 13. Kiran PBS, Manjunath K, Sarkar V (2015) limited power control of a single-stage grid connected photovoltaic system. In: Annual IEEE India conference (INDICON), pp. 1–6 14. Rahim NMA, Quaicoe JE (1996) Analysis and design of a multiple feedback loop control strategy for single-phase voltage-source UPS inverters. IEEE Trans Power Electron 11(4):532– 541

Assessing Small Cross Flow Wind Turbine for Urban Rooftop Power Generation Seralathan Sivamani, R. Hemanth Prasanna, J. Arun, Mikhail Christopher, T. Micha Premkumar, P. Bharath Kumar, Yeswanth Yadav, and V. Hariram

Abstract Horizontal axis wind turbine is not suitable for urban rooftop power generation. The urban wind environment has issues like low wind velocity conditions and turbulence. Crossflow turbine derived from Banki water turbine is used as its structure is simple. Also, it starts at low wind speeds and possess high starting torque. Therefore, the focus of this study is to understand the flow physics of the crossflow wind turbine. Experimental investigations are carried out with wind velocities varying from 4 to 10 m/s. The performance characteristics of crossflow wind turbine are estimated in terms of coefficient of power (C p ), tip speed ratio (λ) and coefficient of torque (C t ). The relationship between time taken to attain the maximum speed at varied wind conditions is also plotted. The results showed that crossflow wind turbine possess a good self-starting ability at relatively low wind speeds with high starting torque. Peak power coefficient, C p,max of 0.0485 is obtained for λ = 0.50. As λ is increased further, the C p value is observed to decrease gradually. C t value is found to decrease almost linearly from a maximum value of 0.096. Polar plot showed that torque output is almost same at all angles which proves the crossflow turbine to be omni-directional. Hence, regardless of the wind directions, crossflow turbine produces same performance making it a suitable choice for the urban rooftop power generation. Keywords Crossflow turbine · Wind energy · Performance · Vertical axis

Nomenclature V

wind speed (m/s).

S. Sivamani (B) · R. Hemanth Prasanna · J. Arun · M. Christopher · T. Micha Premkumar · P. Bharath Kumar · Y. Yadav · V. Hariram School of Mechanical Sciences, Hindustan Institute of Technology and Science, Chennai, Tamil Nadu 603103, India e-mail: [email protected] V. Hariram e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_9

105

106

A D1 H R1 T ρ s λ

S. Sivamani et al.

swept area = A = D1 H, (m2 ). outer diameter of the crossflow wind turbine (m). height of the wind turbine (m) outer radius of the wind turbine = D1/2 (m). torque acting on the wind turbine (Nm). density of air (kg/m3 ). angular velocity (rad/s). Tip speed ratio.

1 Introduction Wind is one among the renewable sources of energy. Wind turbine is a device which converts the wind kinetic energy into mechanical energy. This is in turn gets converted into electrical energy by an alternative current generator. Among the two broad known categories of wind turbines, the horizontal axis wind turbines starts at a typical wind speed of 4.60 m/s onwards. Though horizontal axis wind turbines are commercially more popular, they require a large area and it is not suitable for individual house roof top power generation. On the other hand, vertical axis wind turbines (VAWT) like Savonius runs at low wind speed of 2 m/s, but the efficiency is comparatively less. Among the several types of VAWT, crossflow turbine is recently getting attention among researchers. Interestingly, the crossflow wind turbine is based on the Banki water turbine design. A schematic representation of crossflow wind turbine is shown in Fig. 1a. It has a crossflow runner through which the air passes transversely interacting with the turbine blades both at the entrance and at the exit as depicted in Fig. 1b [1, 2]. It works on the principle of drag force (impulse) where the wind pushes a turbine blade, which in turn rotates the shaft and moves the next blade into position to receive the wind. Generally, it houses about 8--20 curved vertical blades mounted on two horizontal circular disks that are fixed to the shaft. The crossflow

Fig. 1 Crossflow wind turbine

Assessing Small Cross Flow Wind Turbine for Urban Rooftop …

107

turbine has good performance characteristics at low wind speed conditions and is self-starting. It is also simple in design, compact in size, safe to operate and possess a relatively high starting torque as well as power coefficient. A large number of blades also ensure a smooth torque and non-pulsating power. The crossflow wind turbine can also be used for small-scale power generation and for domestic purposes. They can be placed in terraces or in balconies (places with low wind speed and varying wind direction) where other turbines are not favorable. Moreover, it is understood from the existing literatures [3] that VAWT gives a better performance in skewed wind flow conditions which is prevalent in urban wind environment. The crossflow wind turbine design captures the skewed wake of the incoming air flow leading to improved performance and this study is intended to take this phenomenon as an advantage. Diniar Mungil Kurniawati et al. [4] analyzed this type of wind turbine performance with respect to number of blades at low wind speeds (viz., 2–5 m/s). The performance was found to be the best with 16 blades arrangement (C p = 0.21 at λ = 0.59; C t = 0.38 at λ = 0.40). Using crossflow wind turbine with eight blades arrangement, Ayman A Al-Maaitah [5] reported a maximum C p of 0.30 at λ between 0.35 and 0.60. Sivasegaram [6] investigated for the turbine with 2--6 blades at wind speed of 20 m/s and reported the best performance with six blades. Ushiyama et al. [7] studied the turbine with 18, 15 and 12 blades and concluded that maximum efficiency was given by the crossflow wind turbine with 12 blades. Kono et al. [8] conducted numerical as well as experimental investigations on crossflow wind turbine to understand its flow characteristics. Junichiro Fukutomi et al. [9] improved the performance of the turbine from C p,max = 0.098 to C p,max = 0.19 using symmetrical casings. Similar study was also conducted by Toru Shigemitsu et al. [10] using symmetrical casing to improve the performance. In order to overcome the non-uniform wind velocity, crosswind flow turbine was also experimented by keeping it in horizontal position [11]. Hence, on the basis of literature review, it is understood that crossflow wind turbine is a potential one to harness wind energy in urban environments which has issues related with wind velocity namely, turbulence and low wind velocity [3, 12, 13]. Moreover, to make use of the wind energy available at low wind speeds, a new turbine design is to be studied. Therefore, the focus of the study is to analyze the performance parameters of crossflow wind turbine for wind velocity ranging from 4 to 10 m/s. This wind velocity is chosen as it is the prevalent wind velocity range in the urban environment of Chennai city.

2 Experimentation Figure 2 depicts the geometry description of crossflow wind turbine. The design parameters are elaborately discussed in the literature [13]. The outer diameter, D1 is selected as 500 mm. Hence, the radial rim width, as well as inner diameter, D2 is

108

S. Sivamani et al.

Fig. 2 Geometry description

85 and 330 mm respectively. Aperture of the blades, θ b is 29.85° and the number of blades is 20. Blade shape is circular arc with radius Rb = 81.5 mm. As depicted in Fig. 3, the crossflow turbine along with its blades is fabricated using fibre reinforced plastic (FRP). The experimental setup along with crossflow turbine is illustrated in Fig. 4. Open jet wind tunnel equipped with an axial fan of one metre diameter has a speed range from 400 RPM to a maximum speed of 982 RPM with a discharge 11 m3 /s and a motor capacity of 5.5 kW is chosen to create a free stream air flow with varying wind velocity in the laboratory. The speed of axial fan is varied by varying a variable frequency drive. The experimental setup is equipped with sensors along with data acquisition system (National Instruments make and LABVIEW software) interfaced to the computer [14]. A retro-reflective RPM sensor is utilized to measure the turbine rotational speed [15]. Three cup anemometer with a range of 0–60 m/s is used to determine the wind speed [16]. Mechanical type brake drum dynamometer is deployed to apply torque on the shaft of the crossflow turbine [17]. Sushma™ make torque sensor is used to record the torque applied on the turbine’s shaft [18]. The experiment is conducted for a range of wind velocity from 4 to 10 m/s. Data obtained are computed and represented as non dimensional parameters as specified in Eqs. (1), (3) and (4). Tip speed ratio λ = ω R1 /V

(1)

Assessing Small Cross Flow Wind Turbine for Urban Rooftop …

109

Fig. 3 FRP model of crossflow turbine

2π N T 60

(2)

P 0.5ρ AV 3

(3)

Shaft power, P = ωT = Power coefficient C p = Torque coefficient Ct =

T 1 ρ AV 2 D21 2

(4)

3 Results and Discussion The variations between co-efficient of power (C p ) with respect to tip speed ratios are illustrated in Fig. 5. The maximum C p realized for wind velocity 4 m/s and 5 m/s is 0.0365 and 0.04854 respectively. Similarly, for wind velocity 6 m/s, 7 m/s, 8 m/s, 9 m/s and 10 m/s, the maximum C p realized is 0.0437, 0.0421, 0.03826, 0.041, and 0.0437, respectively. As can be seen in Fig. 5, maximum C p realized is at λ = 0.5 for

110

Fig. 4 Experimental setup

Fig. 5 TSR versus C p

S. Sivamani et al.

Assessing Small Cross Flow Wind Turbine for Urban Rooftop …

111

Fig. 6 TSR versus C t

most of the wind velocities. Figure 6 illustrates the changes in coefficient of torque (C t ) with respect to λ at different wind velocities. The maximum C t for the wind velocity 4 m/s is 0.0969. Similarly, for 5 m/s, 6 m/s, 7 m/s, 8 m/s, 9 m/s and 10 m/s, maximum C t is 0.0865, 0.087, 0.0981, 0.0954, 0.096 and 0.0942, respectively. The crossflow wind turbine exhibits a good self-starting ability at relatively low wind speeds with high starting torque. As λ is increased further, C p value is observed to decrease gradually. C t value is found to decrease almost linearly from a maximum value of 0.096. Figure 7 illustrates the changes in torque with respect to angles. As seen in figure, the crossflow turbine produces similar torque for all wind speeds with very little variation. Polar plot showed that torque output is almost same at all angles which proves the turbine to be omni-directional. Figure 8 plots the time taken by crossflow turbine to attain the stable rotational speed under no load conditions for different wind speeds. The crossflow turbine takes 55 s, 58 s, 56 s, 58 s, 60 s, 54 s and 56 s to reach the highest rotational speed for wind speed 4 m/s, 5 m/s, 6 m/s, 7 m/s, 8 m/s, 9 m/s and 10 m/s, respectively. The corresponding rotational speed achieved for these wind speed is 75.681 rpm, 94.444 rpm, 119.422 rpm, 155.342 rpm, 180.213 rpm, 220.522 rpm and 234.636 rpm, respectively. Figure 9 plots the time taken by the crossflow turbine to come to still condition from its maximum rotational speed under no load conditions. The turbine takes 74 s, 82 s, 85 s and 88 s to stop as the turbine rotates at wind speeds of 4 m/s, 5 m/s, 6 m/s and 7 m/s, respectively. Similarly, it takes 90 s, 92 s and 95 s to stop for wind speed of 8, 9 and 10 m/s, respectively.

112

S. Sivamani et al.

Fig. 7 Torque developed with respect to angles

Fig. 8 Time versus RPM for different wind velocities

4 Conclusion Experimental investigations are carried out to understand the flow physics of the crossflow wind turbine with wind velocity varying from 4 to 10 m/s. The crossflow wind turbine exhibits a good self-starting ability at relatively low wind speeds with high starting torque. The performance characteristics are estimated in terms of coefficient of power, tip speed ratio and coefficient of torque. The peak power coefficient, C p,max of 0.0485 is obtained for λ = 0.50. As λ is increased further, C p value is

Assessing Small Cross Flow Wind Turbine for Urban Rooftop …

113

Fig. 9 Time versus RPM for different wind speeds

observed to decrease gradually. C t value is found to decrease almost linearly from a maximum value of 0.096. Polar plot showed that torque output is almost same at all angles which proves the turbine to be omni-directional. Hence, crossflow wind turbine can be utilized for urban rooftop power generation. Moreover, the prototype of this wind turbine can be scaled according to the operational requirements.

References 1. Mockmore CA (1949) Fred Merryfield: The Banki water turbine. Bulletin Series No. 25, Oregon State College, Corvallis, USA 2. Chichkhede S, Verma V, Gaba VK, Bhowmick S (2016) A simulation based study of flow velocities across cross flow turbine at different nozzle openings. Procedia Technol 25:974–981 3. Chong WT, Muzammil WK, Wong KH, Wang CT, Gwani M, Chu YJ, Poh SC (2017) Cross axis wind turbine: pushing the limit of wind turbine technology with complementary design. Applied Energy 207:78–95 4. Kurniawati DM, Tjahjana DDDP, Santoso B (2018) Experimental investigation on performance of crossflow wind turbine as effect of blades number. In: 3rd international conference on industrial, mechanical, electrical and chemical engineering, AIP conference proceedings 1931, pp 030045-1–030045-7 5. Al-Maaitah AA (1993) The design of the Banki wind turbine and its testing in real wind conditions. Renewable Energy 3(6-7):781–786 6. Sivasegaram S (1978) An experimental investigation of a class of resistance-type directionindependent wind turbines. Energy 3:23–30 7. Ushiyama I, Isshiki N, Chai GZ (1990) Experimentally evaluating the design and performance of crossflow wind rotors. In: 1st World renewable energy congress, 10 pages 8. Kono T, Yamagishi A, Kiwata T, Kimura S, Komatsu N (2016) Experimental and numerical investigation on the flow characteristics around a cross-flow wind turbine. Energy Power Eng 8(4):173–182

114

S. Sivamani et al.

9. Fukutomi J, Shigemitsu T, Daito H (2011) Study on performance and flow condition of a crossflow wind turbine with a symmetrical casing. ASME J Fluids Eng 133:051101-1–051101-9 10. Shigemitsu T, Fukutomi J, Takeyama Y (2009) Study on performance improvement of crossflow wind turbine with symmetrical casing. J Environ Eng 4(3):490–501 11. Santoso B, Tjahjana DDDP (2018) The influence of guide vane to the performance of cross-flow wind turbine on waste energy harvesting system. MATEC Web Conf 159(02014) 6 pages 12. Chong WT, Fazlizan A, Poh SC, Pan KC, Hew WP, Hsiao FB (2013) The design, simulation and testing of an urban vertical axis wind turbine with the omni-direction-guide vane. Appl Energy 112:601–609 13. Dragomirescu A (2011) Performance assessment of a small wind turbine with crossflow runner by numerical simulations. Renewable Energy 36:957–965 14. Micha Premkumar T, Seralathan S, Gopalakrishnan R, Mohan T, Hariram V (2018) Experimental data of the study on H-rotor with semi-elliptic shaped bladed vertical axis wind turbine. Data Brief 19:1828–1836 15. Premkumar TM, Sivamani S, Kirthees, E, Hariram V, Mohan T (2018) Data set on the experimental investigations of a helical Savonius style VAWT with and without end plates. Data Brief 19:1925–1932 16. Sarath Kumar R, Micha Premkumar T, Seralathan S, Hariram V (2018) Numerical analysis of different blade shapes of a Savonius style vertical axis wind turbine. Int J Renew Energy Res 8(3:1657–1666 17. Seralathan S, Micha Premkumar T., Mohammed S, Mohan T, Hariram V (2017) Experimental data on load test and performance parameters of a LENZ type vertical axis wind turbine in open environment condition. Data Brief 15:1035–1042 18. Seralathan S, Micha Premkumar T, Vinod Kumar D, Kiran Kumar Reddy V, Dilip Reddy K, Dinesh Reddy K, Hariram V (2019) Experimental data on analysis of a horizontal axis small wind turbine with blade tip power system using permanent magnetic generator 23 Article ID 103716, 9 pages

A Novel Power Electronic-Based Maximum Power Point Tracking Technique for Solar PV Applications D. Ravi Kishore , T. Vijay Muni , and K. S. Srikanth

Abstract The conductance of solar cell was nonlinearly with various atmospheric changes, which had resulted in false MPP. To overcome this problem, the proposed system is educated with the neural network. Although incremental conductance can grant marginally better overall performance in case of rapidly various atmospheric conditions, the accelerated complexity of the algorithm will require larger steeply priced hardware and therefore can also have an advantage over MPPT solely in massive PV arrays. The shortcoming of the fuzzy logic approach is lack of expertise of the obligation cycle variation, which results in suited accuracy level with terrible dynamic characteristics. To overcome the problem, PWM technique is applied, and the corresponding duty cycle is varied to trigger the inverter. Keywords Incremental conductance · MPPT · Solar power efficiency

1 Introduction The performance and efficiency of medium- and large-scale photovoltaic (PV) plant mainly depends on power conversion process through power electronic converter interfaced between PV plant and the distribution grid or load. The power conversion system used in the existing solar PV (SPV) plant involves high cost and reduced efficiency, due to more than one power conversion stage [1]. The existing single-stage power conversion system has low voltage gain and common mode leakage current problem and requires more number of PV panels to get the desired output. Moreover D. Ravi Kishore Department of EEE, Godavari Institute of Engineering & Technology (A), Rajahmundry, India T. Vijay Muni (B) Department of Eelectrical and Electronics Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India e-mail: [email protected] K. S. Srikanth Department of Eelectrical and Electronics Engineering, Amalapuram Institute of Management Sciences & College of Engineering, Mummidivaram, AP, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_10

115

116

D. Ravi Kishore et al.

the output power from PV plant is intermittent which affects the reliability of supply and performance of power converters [2]. The fluctuating power leads to the harmonic injection into the grid and affects real and reactive power flow resulting in the reduction of efficiency of the power conversion system. To overcome the above problems and to improve the performance of PV power conversion system, this research work develops a multistage DRSS and SDC algorithms to enhance the performance of PV systems This research work also investigates MPPT for PV system. MPPT controller is used to extract the maximum available power from PV plant, so that efficiency of PV system can be increased. In this research work, power electronic-based MPPT controller is used [3–5]. The conventional MPPT algorithms have large oscillations due to continuous perturbation, sluggish response, and low efficiency due to ripple in the output voltage and require more number of sensors during the sudden changes in irradiance condition. To minimize the above shortcomings, a new MPPT algorithm has been developed with slope detection method and variable step perturbation and observation algorithms [6]. The performance of the proposed algorithm is investigated and compared with variable step incremental conductance and based on step change in irradiance condition, efficiency, ripple, and response time.

2 Problem Statement and Objectives of Research Work The solar-powered system introduces a variety of issues based on climate conditions. These problem formulations are listed as follows: (i) The conductance of solar cell was nonlinearly fluctuating with climatic changes, which had produced in false maximum power point. (ii) The existing incremental conductance methods are not able to provide better production in case of quickly varying climatic circumstances. (iii) The Solar panel power generation efficiency diminishes based on the light intensity, temperature, and dust. (iv) The previous MPPT techniques increase the complexity of the algorithm which will require more expensive hardware such as large PV arrays. (v) The shortcoming of the MPPT approach is the confusion of the duty cycle difference, which appears inadequate accuracy level with poor dynamic features [7]. The idea of the consideration is to achieve the maximum power applying PV panel, which can be performed in different ways. The following are the primary objectives of the research work. (i) The main objective of this work to design advanced maximum power point (MPPT) tracking algorithms to enhance the solar power generation, (ii) to implement dynamic soft switching logic controllers for receive maximum power point from solar panel, (iii) to use dynamic soft switching controllers for exploring the unique models of photovoltaic power generation, (iv) to use different types of controllers for solar power generation during quick changes in sun irradiance for different environmental situation, and (v) to enhance the solar power generation using MPPT techniques and also reduce the switching losses as well as lower order harmonics.

A Novel Power Electronic-Based Maximum Power Point …

117

3 Proposed Methodology The proposed strategy is to push the framework to execute remarkable systems to expand the productivity of the solar power generation like self-cleaning, auto-cooling, and tracking system that has been designed. PV framework cannot appear as a steady DC source since its output control is moved relying on nature, temperature, and light intensity. Maximum power point tracking is utilized to track the most exceptional power in the photovoltaic structure. The feasibility of solar energy relies on both the MPPT control frameworks and the MPPT circuit [8]. The MPPT control incremental conductance procedure is connected to the DC–DC converter, which is utilized as the MPPT circuit. In this, proposed work unveils a possible approach managed out how to enhance the capacity of solar power generation in various structures by the use of reflected mirrors, auto-tidy cleaning, and customized cooling framework, and this structure is produced utilizing locally available raw materials to influence it to financially efficient one [9]. Incremental conductance structure is utilized to track the MPP under low irradiance (Fig. 1). The proposed work is apportioned into four zones 1. self-cleaning, 2. programmed cooling, 3. mirror reflection, and 4. incremental conductance strategy.

3.1 Self Cleaning In self-cleaning process, two DC motors of 1000 rpm are joined together for cleaning with. The cleaning wiper is utilized for cleaning the dust. The microcontroller is used to operate the DC motors based on the RTC time value. The general work process outline of the self-cleaning system appears in Fig. 2. Based on the real-time clock

Fig. 1 Block diagram of the proposed system

118

D. Ravi Kishore et al.

Fig. 2 Self-cleaning flowchart

parameter, the microcontroller will operate the cleaning wiper for every one hour time gap.

3.2 Automatic Cooling In the automatic cooling system, LM35 temperature sensor is fixed backside of the solar panel. Based on the temperature sensor response, the microcontroller will operate the cooling system automatically. Figure 3 talks about the work process graph for a programmed cooling framework. Absolutely when the temperature value is above 35 °C, the cooling system is switched on automatically with the help of microcontroller. The cooling system has two cooling DC fans and chilling water pump system.

A Novel Power Electronic-Based Maximum Power Point …

119

Fig. 3 Automatic cooling system flow chart

3.3 Mirror Reflection Reflected mirrors are used as a piece of this work to construct the scene photograph voltaic radiation on photovoltaic boards. In early hours and around evening time, the power of sun-oriented radiation is low when appeared differently in relation to night. The intensity of sun-powered radiation will be reached out by perspective about sun radiation to sunlight-based board by reflected mirrors. The configuration of reflected reflect is as showed up in beneath noted in Fig. 4. Fig. 4 Mirror reflection setup

120

D. Ravi Kishore et al.

3.4 Incremental Conductance Algorithm The incremental conductance strategy is utilized to track the MPP under the low irradiance circumstance or when sun radiations are humbler. In incremental conduction technique utilizing the resistor divider circuit, the voltage is lessened, in any case, in light of the way that microcontroller works in the midst of (0–5 V), and voltage given by solar PV panel is more, which drawing in by the microcontroller. The divider circuit has showed up underneath with event of solar panel input. The MPP would along these lines have the ability to be trailed by looking at the fluttering conductance (I/V ) to the incremental conductance (I/V ) that has appeared in the flowchart in Fig. 5.

4 Proposed MPPT Algorithm Incremental conductance (IC) was structured dependent on a comment of P–V property bend. This calculation was planned to defeat some disadvantage of P & O calculation. Steady conductance (IC) attempts to improve the following time and to deliver more prominent quality on a huge illumination adjustments condition. The MPP can be determined by methods for the use of the connection between DI/DV and I/V. On the off chance that DP/DV is negative, then MPPT lies on the correct part of current capacity, and if the MPP is high caliber, the MPPT is on left side. The condition of IC strategy is.

Fig. 5 P–V curve of a solar module

DI DP =1+V DV DV

(1)

DP DI =1+V DV DV

(2)

A Novel Power Electronic-Based Maximum Power Point …

MPP is reached when

DP DV

121

=0 DI I =− DV V

DI > 0 then DV DP > 0 then DV DP = 0 then DV

(3)

V p < VMPP V p < VMPP V p = VMPP .

(4)

The incremental conductance strategy relies upon the way that the slope of the PV bunch control bend is zero at the MPP, positive on the left of the MPP, and negative DP on the right, as given by, DV = 0 at MPP. IC algorithm can be seen on Fig. 6.

Fig. 6 Flowchart of incremental conductance method

122

D. Ravi Kishore et al.

Fig. 7 Increase the solar irradiation

4.1 Increase in Solar Irradiance Level At first, the PV module is respected to work at burden line 2, the photograph voltaic illumination is 0.4 kW/m2 , and the current and voltage of the PV module are Vmpp0.4 and Impp0.4 as demonstrated in Fig. 7. At that point, if the sun oriented light will increment to 1.0 kW/m2 , while the commitment cycle of the DC-DC converter remains unaltered, the running purpose of PV module is at factor D (V1, I1) of burden line 2 which is far away from the MPP of 1.0 kW/m2 . Like the calculation utilized on account of limit in sun-oriented light level, the approximated qualities are substituted to make certain the PV module works near to the new MPP. Be that as it may, the working current, I1 is far away from the speedy circuit present day of 1.0 kW/m2 as demonstrated in Fig. 7. Along these lines, an additional progression is required to ensure the working current of PV module is near to the Isc of new MPP. As demonstrated in Fig. 7, factor E, Voc1.0, and Vmpp0.4 shape a right-calculated square shape.

5 Results and Discussion The proposed IC-based MPPT has been arranged using MATLAB/Simulink. Figure 8 displays the Simulink model of proposed system. Figure 9 demonstrates the duty cycle response of our proposed maximum power point tracking (MPPT) system using SDC control. Figure 10 exhibits the transient responses of the following force bends got from SDC rationale control calculation. As found in the figure, the proposed response is

A Novel Power Electronic-Based Maximum Power Point …

123

Fig. 8 Simulink model

Fig. 9 Duty cycle

generously faster than other normal MPPT while the overshoots of the system are about the same. The above Fig. 11 demonstrates the maximum power point (MPP) response and boost voltage response of our proposed system using sophisticated distribution control (SDC) system. Figure 11 shows the solar panel crack detected simulation

124

D. Ravi Kishore et al.

Fig. 10 Photovoltaic current and voltage

Fig. 11 MPPT voltage and boost voltage

diagram for proposed SDC method. By using SDC method, the solar panel faults are clearly identified.

A Novel Power Electronic-Based Maximum Power Point …

125

6 Conclusion The proposed ICA is appropriate in all range of luminance and temperature range. The analysis of PV panel with a variety of parameters helps in controlling the switching operation with environmental problems as a consequence presenting most reliability in power tracking. High efficiency by 99%, output power by 0.98, switching losses by 4%, output energy by 100.2 Wh, and output voltage by 12 V have been obtained by using the high impedance matching, MPPT’s, variable boosting coefficient, and also it added as the advantages for proposed converter.

References 1. Jyothi VM, Muni TV (2016) An optimal energy management system for PV/battery standalone system. Int J Electri Comput Eng 6(6):2538 2. Vijay Muni T, Priyanka D, Lalitha SVNL (2018) Fast Acting MPPT algorithm for soft switching interleaved boost converter for solar photovoltaic system. J Adv Res Dyn Control Syst 10, 09-Special Issue 3. Vijay Muni T, Lalitha SVNL, Krishna Suma B, Venkateswaramma B A new approach to achieve a fast acting MPPT technique for solar photovoltaic system under fast varying solar radiation. Int J Eng Technol 7(2.20):131–135 4. Vijay Muni T, Lalitha SVNL Power management strategy in solar PV system with battery protection scheme. Int J Inn Technol Exploring Eng 8(6):960–964 5. Vijay Muni T, Lalitha SVNL Fast acting MPPT controller for solar PV with energy management for DC microgrid power management strategy in solar PV system with battery protection scheme. Int J Eng Adv Technol 8(5):1539–1544 6. Zhao X, Li YW, Tian H, Wu X (2016) Energy management strategy of multiple supercapacitors in a DC Microgrid Using. IEEE J Emerg Sel Top Power Electron 4:1174–1185 7. Ac H, Microgrid DC, Xiao H, Luo A, Member S, Shuai Z (2016) An Improved control method for multiple bidirectional power converters in. IEEE Trans Power Electron 7:340–347 8. Meng L, Riva E, Luna A, Dragicevic T, Vasquez JC, Guerrero JM (2016) Microgrid supervisory controllers and energy management systems: A literature review. Renew Sustain Energy Rev 60:1263–1273 9. Boukettaya G, Krichen L (2014) A dynamic power management strategy of a grid connected hybrid generation system using wind, photovoltaic and flywheel energy storage system in residential applications. Energy 71:148–159 10. Ilahi S, Ramaiah M, Vijay Muni T, Naidu K (2018) Study the performance of solar PV array under partial shadow using DC-DC converter. J Adv Res Dyn Control Syst 10(04-Special Issue): 1006–1014 11. Moulali S, Vijay Muni T, Balasubrahmanyam Y, Kesav S (2018) A flying capacitor multilevel topology for PV system with APOD and POD pulse width modulation. J Adv Res Dyn Control Syst 10(02-Special Issue): 96–101 12. Tejasreenu T, Srikanth M, Vijay Muni T (2014) THD reduction and voltage flicker mitigation in power system base on STATCOM. In: IEEE international conference on information communication & embedded systems (ICICES 2014). S.A Engineering College, Chennai 13. Vijay Muni T, Venkata Kishore K, Sesha Reddy N (2014) Voltage flicker mitigation by FACTS sevices. In: IEEE international conference on circuit, power and computing technologies (ICCPCT 2014) 14. Vijay Muni T, Lalitha SVNL, Rajasekhar Reddy B, Shiva Prasad T, Sai Mahesh K (2017) Power management system in PV systems with dual battery. Int J Appl Eng Res 12(1):523–529. ISSN 0973–4562

126

D. Ravi Kishore et al.

15. Vijay Muni T, Sai Sri Vidya G, Rini Susan N, Dynamic modeling of hybrid power system with MPPT under fast varying of solar radiation. Int J Appl Eng Res 12(1):530–537. ISSN 0973–4562 16. Srikanth M, Vijay Muni T, Vishnu Vardhan M, Somesh D (2018) Design and simulation of PV-wind hybrid energy system. J Adv Res Dyn Control Syst 10(04-Special Issue): 999–1005 17. Venkata Kishore K, Vijay Muni T, Bala Krishna P (2019) Fuzzy control based iUPQ controller to improve the network of a grid organization. Int J Modern Trends Sci Technol 5(11):40–44 18. Sudharshan Reddy K, Sai Priyanka A, Dusarlapudi K, Vijay Muni T Fuzzy logic based iUPQC for grid voltage regulation at critical load bus. Int J Inn Technol Expl Eng 8(5):721–725 19. Swapna Sai P, Rajasekhar GG, Vijay Muni T, Sai Chand M Power quality and custom power improvement using UPQC. Int J Eng Technol (UAE) 7(2):41–43

Analysis of an Enhanced Positive Output Super-Lift Luo Converter for Renewable Energy Applications A. Paramasivam, K. B. Bhaskar, N. Madhanakkumar, and C. Vanchinathan

Abstract In this research, an enhanced positive output super-lift Luo (EPOSLL) converter is utilized to track maximum power of photovoltaic (PV) source. Further, the drift avoidance perturb and observe (P&O) maximum power point tracking (MPPT) technique is adopted to track maximum power. Also, the two control strategies are proposed with the help of proportional integral (PI) controller to maintain constant output voltage and output power, respectively. The MATLAB/Simulink software is utilized to simulate the entire research work. Results demonstrate that the proposed converter design not only tracks the maximum power but also it maintains the output voltage and power to a fixed set value. Also, the solar irradiance (G) is varied and it is observed that the output voltage and output power are constant at all time and gain high significance in motor drive applications. Keywords Output voltage · Output power · PI control · Super-lift Luo converter · MPPT

1 Introduction Recent years, DC-DC converters gain more significance due to the huge development in renewable energy-based electricity generation [1]. However, several DCDC converters are available; high power density, high efficiency, system compactness, etc., are the important factors to be considered while designing DC-DC converters. Luo converters are simply a boost-type DC-DC converter which utilizes voltage lifting technique for operation. Further, the Luo converters have less system

A. Paramasivam (B) B. S. Abdur Rahman Crescent Institute of Science and Technology, Chennai 600048, India e-mail: [email protected] K. B. Bhaskar · C. Vanchinathan Adhi College of Engineering and Technology, Chennai 631605, India N. Madhanakkumar Mailam Engineering College, Mailam 604304, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_11

127

128

A. Paramasivam et al.

complexity, high voltage transfer gain, high power transfer efficiency, etc., and make more advantageous than the conventional DC-DC converters [2–6]. Presently, various lifting techniques’ multioutput for Luo converters is developed which makes it suitable for any applications. He and Luo [4] have analyzed various Luo converters such as positive output self-lift and super-lift Luo converter, negative output self-lift Luo converter and positive output re-lift converter. Luo [5] has proposed a new double output Luo converters with voltage lift techniques. Tekade et al. [6] have designed a positive output super-lift Luo converter for PV inverter applications. Nowdays, solar-based electrical energy generation is emerging due to its abundant existence. Further, it is little tedious to extract maximum power from photovoltaic (PV) panels. Several researchers have proposed maximum power point tracking methods such as modified perturb and observe and colony of forging ants [7–15]. To course maximum power point increases the efficient extraction of electrical energy. Therefore, an efficient MPPT technique with Luo converters enhances solar-based electrical energy generation. The objective of this work is to propose a DC-DC converter with an efficient MPPT and to maintain output power and output voltage constant for PV applications.

2 Methodology 2.1 Enhanced Super-Lift Luo Converter Figure 1 shows the circuit diagram for proposed an enhanced positive output self-lift Luo converter. The input of the EPOSLL converter is sourced by photovoltaic source.

Fig. 1 Circuit diagram for proposed enhanced positive output self-lift Luo converter

Analysis of an Enhanced Positive Output Super-Lift Luo Converter …

129

The POSLL converter consists of components such as capacitors (C in , C 1 , C 2 , C 3 , C out ), inductor (L 1 ), MOSFET (S 1 ), and resistive load (Ro ). The drift avoidance P&O MPPT technique is utilized to track maximum power of the connected PV source [8]. For this purpose, the panel voltage and current are continuously tracked according to the drift avoidance P&O algorithm, the duty cycle of the S1 is varied. The proposed EPOSLL converter can operate in two modes, namely constant voltage mode and constant power mode. In constant voltage mode, the output voltage of the EPOSLL converter is taken as a feedback to the PI controller and the controller continuously maintains the output voltage to the constant set values. In constant power mode, the output power of the EPOSLL converter is taken as a feedback to the PI controller and the controller continuously maintains the output power to the constant set values.

2.2 Closed-Loop Control Using PI Controller In this work, the PI controller is used to maintain the output voltage and power of the proposed EPOSLL converter. Figures 2 and 3 show the block diagram of PI control scheme to maintain output voltage and output power to the set value. PI controller forms the control ui (t) from the error signal ei (t) between desired or expected output and actual output. The control signal ui (t) can be expressed as: ⎡ 1 u i (t) = K pi ⎣ei (t) + Tpi

t

⎤ e(τ )dτ ⎦

0

where K pi and T pi are PI controller settings (Table 1). Fig. 2 Block diagram of PI control for proposed EPOSLL converter with constant output voltage

(1)

130

A. Paramasivam et al.

Fig. 3 Block diagram of PI control for proposed EPOSLL converter with constant output power

Table 1 Design parameters of EPOSLL converter

Parameters

Variables

Values of EPOSLL converter

Magnetizing inductance (µH)

L

120e−3 µH

Capacitors (µF)

C in , C 1 , C 2 , C 3 , C out

100, 100, 200, 100, 1000 µF

Switching frequency (KHz)

fs

10 Khz

Load resistance ()

Ro

3 Results and Discussion Figure 4a–c shows the output parameters such as output voltage, output current, and power of the EPOSLL converter under constant voltage mode. It is observed that the output voltage of the EPOSLL converter is constantly maintained at 36 V. Further, the value of load resistance is 40  and corresponding output current is 0.9 A. Also, the derived output power from the proposed EPOSLL converter is 32.4 W. Figure 5a, b shows the output response of EPOSLL converter such as output voltage and output current at two different solar irradiance (G) under constant voltage mode condition. The solar irradiance of the PV source is reduced from 600 to 400 W/m2 and it is seen that there is a sudden decrease in both output voltage and output current for very less period of time. Since the proposed EPOSLL converter is controlled by closed-loop PI control, the output voltage and output current are maintained constant and improve the reliability of the EPOSLL converter. The values of these voltage and current sag increase with increase in solar irradiance (G). Table 2 presents the output response corresponding to disturbances in solar irradiance (G). It is observed that the output set voltage of the EPOSLL converter is 36 V at all conditions. Also, the PI controller maintains the output voltage to 36 V at all conditions. Further, it is seen that the sudden decrease in solar irradiance (G) from

Analysis of an Enhanced Positive Output Super-Lift Luo Converter … 40

Output Voltage (Volts)

Fig. 4 a–c Output parameters of EPOSLL converter at constant voltage mode

131

30

20

10

0

0

2

4

6

8

10

8

10

Time (Seconds)

(a)

Ouput Current (Amps)

1 0.8 0.6 0.4 0.2 0

0

2

6

4

Time (Seconds)

(b)

Output Power (W)

40

30

20

10

0

0

2

4

6

8

10

Time (Seconds)

(c)

600 to 500 (G/m2 ) introduces 0.1 V output voltage sag. Also, if the G is decreased from 600 to 400 (G/m2 ), then the 0.3 V output voltage sag is introduced. Similarly, if the G is decreased from 600 to 300 (G/m2 ), then the 0.6v output voltage sag is introduced. Figure 6a–c shows the output parameters such as output voltage, output current, and output power of the EPOSLL converter under constant power mode. It is observed that the output power of the EPOSLL converter is constantly maintained at 200 W.

A. Paramasivam et al. 0.98

37 36

Output Current (Amps)

G (W/m2)

Output Voltage (V)

132

35 34

600 500 400 0

2

4

6

8

0.96 0.94 0.92 0.9 0.88 0.86 0.84

10

0

2

4

6

Time (Seconds)

Time (Seconds)

(a)

(b)

8

10

Fig. 5 a,b Output response of EPOSLL converter under constant voltage mode at two different solar irradiances

Table 2 Output response corresponding to disturbances in solar irradiance (G/m2 )

Solar irradiance G (W/m2 )

Output set voltage (V)

Output voltage (V)

Output voltage sag (V)

600

36

36

0

500

36

36

0.1

400

36

36

0.3

300

36

36

0.6

Further, the value of load resistance is 40  and corresponding output voltage is 89 V. Also, the derived output current from the proposed EPOSLL converter is 2.25 A. Figure 7a–c shows the output response of EPOSLL converter such as output voltage and output current at two different solar irradiance (G) under constant power mode condition. The solar irradiance of the PV source is reduced from 600 to 400 W/m2 , and it is seen that there is a sudden decrease in both output power for very less period of time. Since the proposed EPOSLL converter is controlled by closed-loop PI control, the output power is maintained constant that improves the reliability of the EPOSLL converter. The value of these power sag increases with increase in solar irradiance (G). Table 3 presents the output response corresponding to disturbances in solar irradiance (G). It is observed that the output set power of the EPOSLL converter is 200 W at all conditions. Also, the PI controller maintains the output power to 200 W at all conditions. Further, it is seen that the sudden decrease in solar irradiance (G) from 600 to 500 (G/m2 ) introduces 0.1 W output power sag. Also, if the G is decreased from 600 to 400 (G/m2 ), then the 0.4 W output power sag is introduced. Similarly, if the G is decreased from 600 to 300 (G/m2 ), then the 1 W output power sag is introduced.

Fig. 6 a–c Output parameters of EPOSLL converter at constant power mode

Output Voltage (Volts)

Analysis of an Enhanced Positive Output Super-Lift Luo Converter …

133

80 60 40 20 0

0

2

4

6

8

10

8

10

8

10

Time (Seconds)

(a) Output Current (Amps)

2.5 2 1.5 1 0.5 0

0

2

4

6

Time (Seconds)

(b)

Output Power (Watts)

250 200 150 100 50 0

0

2

4

6

Time (Seconds)

(c)

4 Conclusion In this work, an enhanced positive output super-lift Luo (EPOSLL) converter was utilized with various functionalities such as maximum power point tracking, constant output voltage, and constant output power. Also, these functionalities were achieved with drift avoidance P&O MPPT technique and two control strategies by PI controller to maintain constant output voltage and output power, respectively. Further, with the

200 100 0

G (W/m2)

650 600 550 500 450

0

2

4

6

8

10

Time (Seconds)

(a)

Output Voltage (Volts)

80 60 40 20 0

0

2

6

4

8

10

Time (Sec)

(b) 2.5

Output Current (Amps)

Fig. 7 a, b Output response of EPOSLL converter under constant voltage mode at two different solar irradiance

A. Paramasivam et al. Output Power (W)

134

2 1.5 1 0.5 0

0

2

6

4

Time (Seconds)

(c)

8

10

Analysis of an Enhanced Positive Output Super-Lift Luo Converter … Table 3 Output power corresponding to disturbances in solar irradiance (G/m2 )

135

Solar irradiance G (W/m2 )

Output set power (W)

Output power (W)

Output power sag (W)

600

200

200

0

500

200

200

0.1

400

200

200

0.4

300

200

200

1

help of MATLAB/Simulink software, the simulation work was carried out. Results demonstrate that the proposed converter design is highly significant to course the maximum power and it is capable to maintain constant output voltage and power. Also, the performance was measured by varying solar irradiance (G) and it is observed that the output voltage and output power were maintained constant.

References 1. Rajini V, Paramasivam A (2013) Analyzing wind power potential in Cauvery delta areas for implementation of renewable energy based standalone pumping system for irrigation. IERI Proc 5:153–160 2. Luo FL (2008) Analysis of super-lift Luo-converters with capacitor voltage drop. In: 2008 3rd IEEE conference on industrial electronics and applications, pp 417–422. IEEE 3. Luo FL (1998) Luo-converters, voltage lift technique. In: PESC 98 record. 29th annual IEEE power electronics specialists conference (Cat. No. 98CH36196), vol 2, pp 1783–1789. IEEE 4. He Y, Luo FL (2005) Analysis of Luo converters with voltage-lift circuit. IEE Proc Electric Power Appl 152(5):1239–1252 5. Luo FL (1998) Double output luo-converters-voltage lift technique. In: 1998 international conference on power electronic drives and energy systems for industrial growth. Proceedings, vol 1, pp 342–347. IEEE 6. Tekade Anushka S, Juneja R, Kurwale M, Debre P (2016) Design of positive output super-lift luo boost converter for solar inverter. In: 2016 international conference on energy efficient technologies for sustainability (ICEETS), pp 153–156. IEEE 7. Sundareswaran K, Vigneshkumar V, Sankar P, Simon SP, Srinivasa Rao Nayak P, Palani S (2015) Development of an improved P&O algorithm assisted through a colony of foraging ants for MPPT in PV system. IEEE Trans Ind Inf 12(1), 187–200 8. Killi M, Samanta S (2015) Modified perturb and observe MPPT algorithm for drift avoidance in photovoltaic systems. IEEE Trans Industr Electron 62(9):5549–5559 9. Elgendy MA, Atkinson DJ, Zahawi B (2016) Experimental investigation of the incremental conductance maximum power point tracking algorithm at high perturbation rates. IET Renew Power Gener 10(2):133–139 10. Messalti S., Harrag AG, Loukriz AE (2015) A new neural networks MPPT controller for PV systems. In: IREC2015 the sixth international renewable energy congress, pp 1–6. IEEE 11. Pradhan R, Subudhi B (2015) Double integral sliding mode MPPT control of a photovoltaic system. IEEE Trans Control Syst Technol 24(1):285–292 12. Mohd A, Jain S, Agarwal V(2019) A time-based global maximum power point tracking technique for PV system. IEEE Trans Power Electron 13. Yaylacı EK (2019) Discrete-time integral terminal sliding mode based maximum power point controller for the PMSG-based wind energy system. IET Power Electron 12(14):3688–3696

136

A. Paramasivam et al.

14. Cardoso RdeB, da Silva ERC, Fernandes DA Multilevel-boost-converter-neutral-pointclamped-inverter photovoltaic system with MPPT based on fibonacci search. In: 2019 IEEE energy conversion congress and exposition (ECCE), pp 4597–4604. IEEE 15. Zushi Y, Nagai Y, Tanimoto T, Tomita Y Distributed maximum power point tracking control under sudden partial shade using an isolated modular boost converter for automotive application. In: 2019 IEEE energy conversion congress and exposition (ECCE), pp 3407–3412. IEEE

Power Quality Enhancement of DC Micro-grid Using DC Electric Spring A. G. Anu, R. Hari Kumar, and S. Ushakumari

Abstract DC micro-grids are prone to voltage instability and voltage flickers due to the incorporation of renewable sources, nonlinear loads, and occasional faults. DC Electric Spring (DCES) is an innovative technology preferred for mitigating the voltage distortions in power systems. This chapter explores the prospects of DCES in enhancing the power quality of Solar Photovoltaic (SPV) system fed DC microgrid. The test system comprises of PV sources with MPPT using perturb and observe method, battery, DCES, and loads. The analysis is carried out in MATLAB/Simulink with and without DCES. The simulation results reveal the effectiveness of DCES in stabilizing the system bus voltage and extending the life time of battery in the DC micro-grid and thereby reducing the environmental hazards from the end of life disposal of batteries. Keywords DC micro-grid · Power quality · DC electric spring · Smart load · Non-critical load

1 Introduction Environmental concerns, depletion of fossil fuels, and increased load demand have necessitated the increased utilization of renewable energy sources in modern grid, though restrictions put forth by regulations, grid stability, and poor power quality in the system have prevented utilization of intermittent renewable energy sources to its fullest. These hurdles could be overcome by using micro-grid. Generally, micro-grids are classified as DC micro-grid, AC micro-grid, and hybrid AC-DC micro-grid. A. G. Anu (B) · R. H. Kumar (B) College of Engineering Trivandrum, Thiruvananthapuram, Kerala, India e-mail: [email protected] R. H. Kumar e-mail: [email protected] S. Ushakumari TrEST Research Park, Thiruvananthapuram, Kerala, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_12

137

138

A. G. Anu et al.

The DC micro-grid concept was initially introduced by Thomas Alva Edison way back in 1882, but it failed to gain popularity due to high transmission losses at low voltages and lost its ground to more efficient alternating current systems. The drawbacks faced by the early DC systems have been overcome with the recent advancements in power electronics. The renewable energy sources, modern electronic loads, and storage devices, all being inherently DC in nature, eliminates the need for additional power conversion stages, reactive power compensation, and synchronization. This in turn makes the DC grid a more efficient alternative, especially to small scale residential applications. Moreover, this could also be a promising option to extend a sustainable, economical, and reliable solution for remote electrification projects in rural areas. All these put together makes the DC micro-grid one of the most preferred alternative for sustainable energy system. Even though DC grids have aforementioned advantages, they are prone to voltage instability and voltage flickers due to the incorporation of nonlinear loads, non-dispatchable renewable sources, and occasional faults. The biggest challenge faced by DC micro-grid is achieving dynamic voltage regulation due to the fact that modern electronic loads are hypersensitive to voltage variation and also the stability of DC micro-grid is dependent on power balance which in turn is reflected by constant bus voltage. Power generation independent of load demand is the strategy followed in conventional power systems. The real-time determination of power generation at any instant is impractical due to the varying nature of photovoltaic and wind power generation. The stability in any power system mainly relies on the power supply--demand balance. So the existing control strategy has to be transformed to generate power in accordance with load demand. Instantaneous energy balance in real time is made viable by the usage of batteries. Environmental hazards from the disposal of defunct batteries and its prohibitive costs limit its usage. Reducing the size of battery storage elements or prolonging the lifetime of batteries would be a smarter alternative to surpass the aforementioned shortcomings. Electric spring (ES) has been propounded as a new concept to improve the stability of modern grid with non-dispatchable renewable energy sources having fluctuating characteristics. Previous research was predominantly focused on the applicability of electric springs in AC power systems. In [1], AC electric springs have been proved to assist in achieving system voltage stabilization even in extreme fluctuations caused by the intermittent nature of wind power. In [2], AC electric springs have been efficiently proved to reduce battery storage requirement. In [3], the voltage and frequency control of AC electric springs have been proved experimentally. A comparative study between STATCOM and AC electric spring in [4] has proved that electric springs are more efficient. The features of AC electric springs in dynamic voltage regulation, nullifying the demand supply gap and power quality improvement are well proved in literature. But the research on electric spring in the DC systems is minimal. In [5] the electric spring was initially conceptualized as a means for solving issues of DC systems. The advantages and suitability of selecting low voltage DC for residential applications were explored in [6, 7]. In [8], the performance of series-type ES versus shunt-type ES has been analyzed. This work focusses the feasibility of applying the concept of ES on low voltage DC micro-grid by conducting a simulation study in a

Power Quality Enhancement of DC Micro-grid …

139

SPV fed DC micro-grid test system for dynamic voltage regulation and enhancement of the life time of battery storage. This chapter is outlined in five sections. The concept of electric spring, design, its operating modes, and control logic are briefed in Sect. 2. Section 3 deals with the block diagram representation and specifications of SPV fed DC micro-grid test system. Section 4 covers discussions on simulation results. The conclusions are outlined in Sect. 5.

2 DC Electric Spring The underlying principle of mechanical springs put forth by Robert Hooke is analogous to the working of electric springs. Mechanical springs are devices designed to produce force by varying the displacement from its neutral position. It is capable of storing mechanical energy when pressed and releasing equivalent energy when stretched. Electric spring concept was introduced as a method to boost or buck voltages depending on system conditions. In control system engineering, it is a well proved fact that force and voltage are analogous quantities. Due to the above said operational analogies with mechanical spring, it is termed as electric spring. In electric spring theory, on the basis of voltage tolerance limit, the electrical loads are categorized into critical loads (CL) and non-critical loads (NCL). CL needs a constant DC voltage whereas NCL can tolerate wider DC voltage variation within an acceptable range. Devices such as TV, PC, and data servers are examples of CL and equipments like air conditioners, electric water heater, washing machine, etc., come under NCL category. Depending on the operating principle, the DCES are divided into two types, namely shunt DCES and series DCES. Figure 1 represents the basic circuit diagram of series and shunt DCES. The series DCES modelled to act as a controllable voltage source which is connected in series with a NCL forming a smart load [8, 9]. When the bus voltage exceeds the reference voltage, it absorbs more current to boost the NCL voltage and when the bus voltage drops below the reference voltage, it reduces the smart

Fig. 1 Basic circuit of a series DCES b shunt DCES

140

A. G. Anu et al.

load current to decrease the voltage across NCL. The shunt DCES operates as a controllable current source is connected directly to the DC bus and is not related with any NCL [8]. When the bus voltage crosses the reference voltage, the DCES absorbs current from the bus and when the bus voltage is lower than reference it injects current to the bus. Basically, both types of DCES are DCDC converters with storage devices. It also incorporates a voltage controller circuit for achieving DC bus voltage regulation. But in series DCES, bus voltage regulation is obtained by surrendering the voltage quality of NCL whereas shunt DCES operates to maintain a steady voltage across all the loads connected to the bus [10]. Hence, in this work, the shunt-type configuration is selected for analysis. The basic structure of DCES is represented in Fig. 2. The DCES comprises a battery system and a DC-DC converter which is bidirectional in nature. When the voltage on the supply side varies, the DCES operates to inject/absorb a suitable voltage to electric spring making the NCL to adapt automatically to maintain power balance. Thus, a constant voltage is maintained across NCL. In the future, DCES could possibly support in reducing the storage requirements of battery in smart grid. The charging and the discharging cycles of battery can be minimized by the usage of DCES as it also operates to maintain the power balance.

Fig. 2 Basic configuration of DCES

Power Quality Enhancement of DC Micro-grid …

141

2.1 Design of DCES The bidirectional topology of buck-boost converter is derived by the introduction of the bidirectional conducting switch. During boost mode of operation, switch Q1 conducts at the required duty cycle and the complimentary mechanism between switches keeps the switch Q2 open. Similarly, during the buck mode of operation, the switch Q2 is made to conduct at required duty cycle keeping the switch Q1 in off condition. A small dead time is incorporated for eliminating the cross conductance through two switches and the converter output capacitance during mode transition.

2.2 Buck Mode When Q1 is on  Vbatt − Ves DT L1   Vbatt − Ves DT = L1 

VL1 = i L1on

(1) (2)

When Q1 is off VL1 = −Ves = L  i L1off =

di L1 dt

−Ves (1 − D)T L1

(3)  (4)

Applying Volt-sec balance principle in (3) and (4) i L1on + i L1off = 0

(5)

Ves =D Vbatt

(6)

i L1min =

Ves (1 − D)Ves − R 2L 1 f

(7)

i L1max =

Ves (1 − D)Ves + R 2L 1 f

(8)

From ripple current waveform,

142

A. G. Anu et al.

Solving (7) and (8) L1 =

(1 − D)Ves f i L1

(9)

where f = switching frequency, D = duty cycle and V es = output voltage of DCES. Boost mode When Q2 is ON VL1 = L

di L1 = Vbatt dt

(10)

DT Vbatt L

(11)

i L1on = During turn off

di L1 = Vbatt − Ves dt   (Vbatt − Ves )(1 − D)T = L1 L

i L1off

(12) (13)

Applying Volt-sec balance Ves 1 = Vbatt 1− D

(14)

The output voltage ripple for converter is computed from capacitor current waveform. Q = Ies DT = C f 1 Ves Cf1 = Cf2 =

Ies D f Ves

(15) (16)

2.3 Operating Modes of DCES The four operating modes of DCES are. 1. Boosting Discharge Mode (BD): This mode gets activated when the supply voltage falls below the desired value. In this mode, a discharge from the battery would help boost up the bus voltage to restore system stability.

Power Quality Enhancement of DC Micro-grid …

143

2. Boosting Charge Mode (BC): This mode is activated and operates similar to that of BD, however, the battery also gets charged by storing the surplus power. 3. Suppressing Discharge Mode (SD): This mode gets activated when the bus voltage rises above the desired value. In this case, the battery would discharge and transfer excess power to the bus which would in turn step down the voltage at the bus. 4. Suppressing Charge Mode (SC): The operation in SC mode is similar to the SD, but instead of delivering the excess power to bus, the battery of DCES stores the power [8, 9]. 5. The operating modes of series DCES include BD mode, BC mode, and SC mode whereas shunt DCES operates only in BD mode and SD mode.

2.4 Control Logic of DCES Control logic of DCES is as shown in Fig. 3. The reference voltage is compared with bus voltage producing an error signal. The error signal thus produced is then compared with a triangular waveform generating switching pulses for buck/boost operation. A selector switch is used to select buck/boost operation by comparing the reference voltage and bus voltage. When bus voltage drops below reference voltage, boost operation is initiated and when bus voltage is greater than reference voltage, buck operation is selected and hence DC bus voltage is regulated at a constant value.

Fig. 3 Control logic of DCES

144

A. G. Anu et al.

Fig. 4 DC micro-grid test setup

3 DC Micro-grid Test System An 800 W, 48 V SPV fed DC micro-grid test setup depicted in Fig. 4, is designed to evaluate the performance of DCES. The test system comprises of PV sources with MPPT using perturb and observe method, battery, DCES, and loads. The sources and loads are connected via a DC/DC converter. The test setup also consists of a battery which assist the loads during source intermittencies. Hence, the entire setup of source, DCES, and the battery is likely to provide the power requirements of all loads during intermittencies in the generation side.

4 Simulation Results and Discussion The DC micro-grid system shown in Fig. 4 is simulated in MATLAB/Simulink environment. The specifications are outlined in Table 1. Table 1 Specifications of test system

Components Specifications PV1, PV2

V oc = 80 V, I sc = 8.5 A, N s = 140, 1000 W/m2

Battery

48 V, 26 Ah

NCL 1

48 V, 400 W

NCL 2

48 V, 100 W

CL

48 V, 300 W

DCES

cf1 = cf2 = 1 µF, L 1 = 7 m H

Power Quality Enhancement of DC Micro-grid …

145

4.1 Micro-grid Without DCES Figures 5, 6, and 7 show the waveforms of output voltage, currents from PV1, PV2, battery, and load current variations with varying irradiance in PV1 and constant irradiance in PV2 without DCES. The waveforms reveal that the battery is trying to achieve the DC output voltage stabilization but it takes large current from battery to stabilize the output voltage. At t = 6 s, one of the NCL is removed. The simulation result shows that the current drawn from the battery after 6 s reduces and gains voltage stabilization. But the utilization of battery storage to meet the supply-demand balance is an expensive method. Hence, some alternative economic solutions be explored to increase the life time of battery to reduce environmental pollutions resulting from disposal of defunct batteries.

Fig. 5 Output voltage variation without DCES

Fig. 6 Current variations from PV1, PV2, and battery without DCES

146

A. G. Anu et al.

Fig. 7 Load current variation without DCES

4.2 Micro-grid with DCES The waveforms of output voltage, load currents, and current from PV1, PV2 and battery, DCES with varying irradiance in PV1 and constant irradiance in PV2 with DCES are shown in Figs. 8, 9, 10 and 11. The waveforms reveal the fact that by using DCES, the current drawn from the battery is drastically reduced to a lower value compared to the case with PV and battery. In this case, the load current requirement is mainly met from DCES and the remaining only be supplied by battery. The simulation results also prove the effectiveness of DCES in regulating constant bus voltage when one of the NCL is removed at t = 6 s. Thus, it proves to be a very effective method to save capital and maintenance costs incurred for the battery system. The drastic reduction in current drawn from the battery clearly indicates that the operation of battery is relieved from overstress and thereby the life time of battery can be increased

Fig. 8 Output Voltage variation with DCES

Power Quality Enhancement of DC Micro-grid …

147

Fig. 9 Load current variations with DCES

Fig. 10 Current variations from PV1, PV2

by the inclusion of DCES in addition to voltage regulation. The usage of shunt DCES configuration also assures the voltage stability of all loads.

5 Conclusions With the growing interest toward the usage of electronic loads, renewable generations and electric vehicles, the DC micro-grids are becoming a promising option for modern power delivery. DC micro-grids are prone to voltage instability and voltage flickers due to the incorporation of renewable sources, nonlinear loads, and occasional faults. The results imply that the DCES stabilizes the bus voltage during varying

148

A. G. Anu et al.

Fig. 11 Current variations from battery and DCES

irradiance and load rejection conditions. Thus, the simulation results substantiate the usage of DCES in enhancing the power quality of the emerging DC micro-grids.

References 1. Hui SY, Lee CK, Wu FF (2012) Electric springs—a new smart grid. IEEE Trans Smart Grid 3(3):1552–1561 2. Lee CK, Hui SY (2013) Reduction of energy storage requirements in future smart grid using electric springs. IEEE Trans Smart Grid 3(3):1282–1284 3. Chaudhuri NR, Lee CK, Chaudhuri B, Hui SY (2014) Dynamic modelling of electric springs. IEEE Trans Smart Grid 5(5):2450–2458 4. Luo X, Akhtar Z, Lee CK, Chaudhuri B, Tan SC, Hui SY (2015) Distributed voltage control with electric springs: comparison with STATCOM. IEEE Trans Smart Grid 209–219 5. Mok KT, Wang MH, Tan SC, Hui SYR (2017) DC electric springs a technology for stabilizing DC power distribution systems. IEEE Trans Power Electron 2(2):1088–1105 6. Dragicevic T, Lu X, Vasquez JC, Guerrero JM (2016) DC grids part II: a review of power architectures, applications, and standardization issues. IEEE Trans Power Electron 31(5):3528– 3549. 7. Sannino A, Postiglione G, Bollen MHJ (2003) Feasibility of a DC network for commercial facilities. IEEE Trans Ind Appl 39(5):14991507 8. Wang MH, Mok KT, Tan SC, Hui SYR (2015) Series and shunt DC electric springs. In: 2015 IEEE energy conversion congress and exposition (ECCE), Montreal, QC 9. Hashem RA et al (2018) Design of an electric spring for power quality improvement in PV based DC grid. In: 2018 IEEE symposium on computer applications and industrial electronics (ISCAIE) 10. Yang T, Mok KT, Tan SC, Hui SYR (2015) Control of electric springs with coordinated battery management. In: Proceedings of the IEEE energy conversion congress and exposition (ECCE), vol 20–24. Montreal, QC, Canada, pp 4740–4746

Mathematical Modelling of Embedded Switched-Inductor Z-Source Inverter for Photovoltaic Energy Conversion T. Divya

and R. Ramaprabha

Abstract Z-source inverters provide single-stage power conversion for photovoltaic (PV) interface as it does the job of boosting and DC-AC conversion. The topology presented here is derived by fusing the switched-inductor cell (SL) in an embedded switched Z-source inverter which eliminates the problem of inverter leg short circuit (SC). Its output voltage varies over a wide range without any requirement of a time delay in turning on the power switches. This inverter generates high gain factor for the same structural elements in comparison with other topologies and is expected to give continuous input current, and hence, it is more suitable for PV and fuel cell interface. The mathematical model of ESI-ZSI along with PV array is presented in this chapter. The advantageousness of the ESI-ZSI inverter over few basic inverters will be presented by comparing the parameters of the similar existing topologies reported in the literature. Keywords Z-source inverter · Modelling · PV module · Incremental conductance

1 Introduction Energy conservation has been a central topic for many government schemes owing to the decrease in conventional energy resources with the rapid increase in energy demand. First preference is given to solar energy or in other words, photovoltaic (PV) energy, because of its abundance in nature, green, and inexhaustible [1, 2]. Existing PV panels have a lower and varying range of input voltage, and thus, the energy converted by photovoltaic panels, intended for any specific application, should be conditioned by a suitable inverter for further use. Also in many applications, the single-stage architecture is often preferred to realize PV interfaced applications due to its considerable reduction in power loss. Such conditions demand for a buck-boost type inverters, and one such inverter to convert and boost the PV voltage for required applications is presented in this paper with PV interfacing. T. Divya · R. Ramaprabha (B) SSN College of Engineering, Kalavakkam 603110, TN, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_13

149

150

T. Divya and R. Ramaprabha

Fig. 1 Impedance source inverter

Conventional VSIs and CSIs experience similar problem such as (1) works only in either buck or boost mode, (2) high harmonics and electromagnetic inference (EMI), and (3) high voltage stress of input range [3–5]. New and combination of topologies have been developed to offer a better boost value for nominal duty ratio. Z-source inverter (ZSI) (Fig. 1.) was developed for extending the output range of inverter by providing both buck and boost operation. But a conventional ZSI also experiences several drawbacks; hence, modifications of existing and new topologies based on ZSI have been suggested, allowing an extended range of power conversion applications [6, 7]. Along the way, another novel power converter having the same functionality as ZSI was proposed, switched boost inverter (SBI) [8, 9]. This work focuses on few topological changes to an existing topology to increase its efficiency while retaining its main operational merits. Basically, an embedded Z-source (EZSI) is considered for its reduced capacitor voltage rating. Adding to this, the switched-boosting network replaces the inductive structures to provide a stable inverter structure. Along with the structure model, generalized modelling of a PV array with MPPT will be presented in this paper as an overall system.

2 Embedded Switched-Inductor ZSI A new topology with a boosting network incorporated in an embedded switched boost inverter (ESBI) [10] is presented as in Fig. 3. The network consisting of three diodes and two inductors called the switched-inductor (SL) cell usually connects the source and the inverter network as shown in Fig. 2a, b which shows SL cell with n ≥ two diodes and inductors. The operation of the modified inverter consists of three basic modes based on the operating conditions of the switches and diodes of the network similar to that of ESBI. The shoot-through in the ZSI is a major boosting factor of the topology. Hence, the duty cycle (D) of the inverter is considered to be above 0.5, and DSt —period of shoot-through is considered 0.2. The modes are briefly explained below.

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

151

Fig. 2 a A switched-inductor network. b SL cell with n ≥ 2

Fig. 3 A embedded switched-inductor ZSI (ES-ZSI)

For Mode-1 (M1), shoot-through stage, all switches (S1 –S4 ) of the inverter plus switch (S) are turned on. The diodes are reversed, and thereby, capacitor’s stored energy charges the inductor’s stored energy. The energy stored of the capacitor can be obtained as using KVL, v Da = v Db = −VC ; i c = −2I L

(1)

v D3 = −VC − Vi

(2)

v L = v L1 = v L2 = VC + Vi

(3)

The inductor voltage is

+Vi . At mode 1, the two inductor The inductor current increases by a slope of Vc2L currents supply iDC to the inverter neglecting the ripple. For M1,

i dc = 2I L and vo = 0

(4)

For Mode-2 (M2), except for S1 and S4 , other switches are turned off for normal boosting stage. The capacitor’s stored energy increases while the inductor’s decrease since its voltage goes negative. Since diode Db is forward biased, the capacitor stored energy is given by KVL,

152

T. Divya and R. Ramaprabha

i c = IL − i dc ; VC = vdc The inductor-stored energy is decreased by a slope of equating parameters: v L = vD1 = vD2 =

(5) Vi −Vc , 2L

and Eq. (4) gives its

Vi − Vc 2

(6)

For M2, the maximum output vo = vdc,max

(7)

For Mode-3 (M3), this is similar to M2 with S2 and S3 which are turned on while the rest are off. The output of the inverter is negative in this mode. For M3, the minimum output vo = −vdc,max

(8)

In Fig. 4, GS –GS4 represent the switching signals of the switches S–S4 . The same topology can be improvised by linking n number of SL cells instead of one with just 2Ls and 3Ds. The gate pulses GS –GS4 are provided with a simple sine PWM method provided in [10, 11], where a sine wave envelope and a straight-line envelope are used for inverter switches and shoot-through switch in the circuit.

3 Modelling of the Performance Parameters 3.1 Overall Voltage Gain Voltage gain factor is calculated by means of the inductor voltage balance law, characterized by Eq. (9) over T (the time period of the modified inverter). t VL dt = 0

(9)

0

Substituting from Eq. (3), (6) in above equation, we get 

D St T 0

 (Vc + Vi )dt +

T D St T

(

Vi − Vc )dt 2

The average capacitor voltage can be obtained on simplifying Eq. (10),

(10)

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

153

Fig. 4 Waveforms of ESI-ZSI

Vc =

1 + D St Vi 1 − 3D St

(11)

From Eq. (5), vdc,max = Vc =

1 + D St Vi = BVi 1 − 3D St

(12)

where B is boosting factor, consequently the peak values of the load voltage are given as vo,max = −vo,min = vdc,max =

1 + D St Vi 1 − 3D St

(13)

154

T. Divya and R. Ramaprabha

3.2 Current Ripple of Inductor Inductor current ripple is derived from the inductor voltage vL = L didtL , and Eq. 3 is represented in Eq. 14, i L =

Vc + Vi D St T 2L

(14)

From Eqs. 11 and 14, |i L | =

D St (1 − D St ) Vi L f s (1 − 3D St )

(15)

From Eq. 15, it can be computed that switching frequency, and the inductance is inversely proportional to the inductor current ripple.

3.3 Voltage Ripple of Capacitor c Capacitor voltage ripple is derived from the capacitor current ic = C dv , and Eq. 1 dt is represented in Eq. 16,

vc D St T

(16)

D St I L T C

(17)

−2I L = 2C This is rewritten as, |vc | =

From Eqs. (1), (5), (14), and the capacitor current balance law, −2I L D St T + (1 − D St )[I L −

(1 + D St )Vi ]T = 0 R(1 − 3D St )

(18)

The inductor current value can be obtained on simplifying the above equation. IL =

(1 − D St )(1 + D St )Vi R(1 − 3D St )2

(19)

Substitute the value of Eq. 19 in Eq. 17 |vc | =

D St (1 − D St )(1 + D St )Vi RC f s (1 − 3D St )2

(20)

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

155

Similarly, from Eq. 20, it can be computed that capacitor voltage ripple is inversely proportional to the capacitance of the inverter.

4 Comparison of Performance Parameters This section briefs the comparative analysis of the modified topology with basic inverter topologies in literature. The comparison will be based on gain factor (B), ripple of the inductor current (|iL |) and ripple of the capacitor voltage (|vc |). Figure 5 presents the circuit diagram of the embedded switched boost inverter [10]. Increased gain factor will be the primary aim of the topology, and its advantages and disadvantages will be obtained by comparing it with embedded switched boost inverter and ZSI (Fig. 5). The derivations of these parameters have been studied with its working in [3, 10]. The related equations of the proposed inverter along with topologies to be compared obtained from the studies are given in Table 1 From the table, it can be observed that the Dst is the common factor for all the parameters in all these topologies. The performance parameters are calculated theatrically using the above derivations. Tables 2 and 3 show the comparative performances of the inverter topologies based on the parameters calculated mathematically. The calculations from the table yield that the given topology produces an increased gain factor in comparison with the basic topology for lesser number of passive elements. Table 4 also illustrates that increasing the shoot through duty cycle also Fig. 5 Embedded switched boost inverter

Table 1 Comparison of parameters Parameters

ZSI

ESBI

ESI-ZSI

B |iL |

1 1−2D St (1−D St )Vi L f s (1−2D St )

1 1−2D St D St (1−D St )Vi L f s (1−2D St )

1+D St 1−3D St D St (1−D St )Vi L f s (1−3D St )

|v c |

D St (1−D St )Vi RC f s (1−2D St )

D St (1−D St )Vi 2RC f s (1−2D St )2

D St (1−D St )(1+D St )Vi RC f s (1−3D St )2

Passive elements (L + C)

2+2

1+1

2+1

156 Table 2 Comparison of gain factor

Table 3 Comparison of inductor current ripple

Table 4 Comparison of capacitor voltage ripple

T. Divya and R. Ramaprabha Dst

ZSI/ESBI

ESI-ZSI

0.15

1.43

2.09

0.2

1.67

3

0.25

2

5

0.3

2.5

13

Dst

ZSI

ESBI

ESI-ZSI

0.15

2.59

0.39

0.49

0.2

2.84

0.5

0.85

0.25

3.2

0.8

1.6

0.3

3.73

1.12

4.48

Dst

ZSI

ESBI

ESI-ZSI

0.15

0.15

0.06

0.15

0.2

0.22

0.08

0.25

0.25

0.31

0.103

0.44

0.3

0.43

0.135

1.18

increases the ripples along with the gain factor; hence, the Dst is varied from (0.15–0.25)

5 Photovoltaic Model 5.1 Modelling of the Cell The photovoltaic (PV) array provides the source for the designed inverter; the PV cell is modelled from the equation of one-diode model [11] (Fig. 6). The figures show the important elements of the PV model which will be designed based on the following equation. Fig. 6 One-diode model

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

157

where Iph , photocurrent Rs , series resistance (very small—neglected) Rsh , shunt resistance (very large—neglected). The I ph equation for a one-diode model is given below, based on which the modelling of the PV cell will be explained. Iph = [Iscr + K i (Tk − Tref )] ∗

λ 1000

(21)

where I ph , current at 25 °C and 1000 W/m2 I scr , SC current at 25 °C and 1000 W/m2 (2.55 A) K i , SC current/temperature coefficient (0.0017 A/k) T k , Actual temperature (in K) T ref , reference temperature (298 K) λ, irradiation (1000 W/m2 ) Other important factors needed to model the PV cell are the reverse saturation current (I rs ) and saturation current (I O ), and the equation for the factors is be given by the equation below Iscr   Ir s =   oc exp Nqs kVAT −1

(22)

where q, electron charge (1.6 × 10–19 C) V oc , open-circuit voltage (21.24 V) N S , no. of cells in series (36) K, Boltzman constant (1.3805 × 10–23 J/K) A, ideality factor (1.6)  IO = Irs

T Tr

3

 exp



 q ∗ Ego 1 1 − Ak Tr T

(23)

where T, operating temperature (K) E go , band-gap energy (1.1 eV @ 25 °C) From the above factors, the output current of the one-diode PV model is given by equation  IPV = Np ∗ Iph −Np ∗ IO



 q ∗ (VPV + IPV Rs ) VPV + (IPV RS ) exp −1 − (24) NS Ak RSh

158

T. Divya and R. Ramaprabha

Fig. 7 PV characteristics

For simplification, N s is considered as 36, and N p is considered to be 1 for simplification. Since we have considered the value of Rsh to be high, the above I PV will change as follows, 

 q ∗ (VPV + IPV Rs ) −1 IPV = Iph −I O exp N S Ak

(25)

The output voltage and current are measured to plot the characteristics of the designed PV; Fig. 7 shows the PV characteristics (I–V and P–V).

5.2 PV Sizing The simulation is designed a source power of (110–120) W, and hence, a PV module with maximum output voltage of 16.54 V and output current of 2.25 A is considered. The model of the PV was referenced from the Solkar 36WP PV module for which the data sheet is also given in [11]. As per our requirement, three PV modules are series connected to form the PV array with a maximum power point voltage of 49.6 V and power of 112 V (3 × 1 array). Figure 13 shows the series connected PV modules used in the work.

5.3 MPPT—Incremental Conductance (InC) Maximum power point tracking is employed to track the maximum power point for the designed PV array. In this work, incremental conductance (InC) method is used for optimum usage of the PV array. The InC provides maximum power by tracking its incremental change in the PV array power using change in its current and voltage [12]. Figure 8 shows the flowchart for InC based on which a MATLAB code is written.

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

159

Fig. 8 InC flowchart

6 Simulation Results 6.1 PV Array Based on the above equations, the following PV module is designed, and each parameter are designed into separate block (Figs. 9, 10, 11, 12, and 13).

6.2 ESI-ZSI Model The simulation circuit of embedded switched-inductor ZSI supplied with the designed PV Array and MPP tracking is presented in Fig. 14. Since the output of the inverter may have unwanted frequency waves, a band pass (LC) filter is employed across load to remove the unwanted frequencies and provide an almost sinusoidal output suitable for applications. The component parameters for the simulation carried out are given below (Table 5). From the simulation of proposed inverter using the above parameters, the output waveforms of the embedded switched-inductor ZSI are obtained as given in Fig. 16. The gating signals to be given to the switches are presented in Fig. 15.

160

Fig. 9 I ph block

Fig. 10 I rs block

Fig. 11 I o block

T. Divya and R. Ramaprabha

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

161

Fig. 12 I PV block

Fig. 13 Series PV module

As discussed before, the embedded switched-inductor Z-source inverter operates at higher gain factor compared to its base topologies.

162

T. Divya and R. Ramaprabha

Fig. 14 Simulation circuit of ESI-ZSI with PV Table 5 Simulation parameters

Fig. 15 Pulse for the switches

Components

ESI-ZSI

Input voltage (V in )

40–50 V

Switching frequency (f s )

25 kHz

Inductors (L 1 , L 2 )

1.5 mH

Capacitor (C)

47 µF

Load (R)

55 

Filter capacitor (C f )

10 µF

Filter inductor (L f )

1.2 mH

Mathematical Modelling of Embedded Switched-Inductor Z-Source …

163

Fig. 16 Voltage and current waveform of ESI-ZSI

7 Inference The basic parameters of ESI-ZS inverter were modelled with Dst as a common factor, and hence, Dst plays the main role in modelling an application-specific inverter circuit. Basic parameters were considered for comparison of the inverters among embedded switched-inductor ZSI, with embedded SBI and conventional ZSI which were made in terms of gain factor, current, and voltage ripple. From this comparison table, it is yielded that the designed symmetrical topology is found to be superior, providing a variable gain factor by varying Dst and with the number of SL cell connected. A suitable PV array was considered and designed for the verification of the simulation results. InC MPPT was also implemented with the PV array to provide the maximum power to the inverter. The results of the designed topology was simulated with a simple pulse-width modulated (PWM) generated pulses sourced from a PV array, and the results are provided. With this, the hardware verification of the topology will be carried out in future works.

References 1. Meinhardt M, Cramer G (2000) Past present and future of grid connected photovoltaic and hybrid-power systems. In: IEEE power engineering society meeting. IEEE, Seattle, WA, pp 1283–1288 2. Palma L (2017) Design and sizing of short term energy storage for a PV system. In: 6th international conference on clean electrical power (ICCEP). IEEE, Santa MargheritaLigure, pp 460–465 3. Peng FZ (2002) Z-source inverter. In: IEEE industry applications conference—37th annual meeting. IEEE, Pittsburgh, PA, pp 775–781 4. Thorborg K (1988) Power electronics. Prentice-Hall International (U.K.) Ltd., London 5. Krein PT (1998) Elements of power electronics. Oxford University Press, London

164

T. Divya and R. Ramaprabha

6. Vinnikov D, Roasto I, Strzelecki R, Adamowicz M (2012) Step-up DC/DC converters with cascaded quasi-Z-source network. IEEE Trans Ind Electron 59(10):3727–3736 7. Gajanayake CJ, Luo FL, Gooi HB, So PL, Siow LK (2010) Extended-boost Z-source inverters. IEEE Trans Power Electron 25(10):2642–2652 8. Gao F, Loh PC, Li SD, Blaabjerg F (2011) Asymmetrical and symmetrical embedded Z-source inverters. IET Power Electron 4(2):181–193 9. AddaRavindranath F, Santanu K, Mishra S, Avinash Joshi T (2013) Analysis and PWM control of switched boost inverter. IEEE Trans Ind Electron 60(12):5593–5602 10. EbrahimBabaei F, Elias ShokatiAsl S, Mohsen HasanBabayi T, Laali S (2016) Developed embedded switched-Z-source inverter. IET Power Electron 9(9):1828–1841 11. Pandiarajan N, RanganathMuthu S (2011) Mathematical modeling of photovoltaic module with simulink. In: 1st international conference on electrical energy systems. IEEE, Newport Beach, CA, pp 258–263 12. Kish GJ, Lee JJ, Lehn PWT (2012) Modelling and control of photovoltaic panels utilising the incremental conductance method for maximum power point tracking. IET Renew Power Gener 6(4):259–266.

Hybrid Algorithms to Track Peak Power in Solar PV Array Under All Irradiation Conditions R. Ramaprabha

and S. Malathy

Abstract The nonlinear relationship between the voltage and power of the PV array and the reliance of the electrical characteristics on the environmental conditions makes it essential to operate the PV array at its optimal operating point. The voltagepower characteristic exhibits single peak under homogeneous irradiation conditions and multiple peaks under heterogeneous irradiation conditions. This chapter presents and compares two hybrid MPPT algorithms that can track the optimal power point under all irradiation conditions. Both the algorithms are based on divide and discard strategy, and hence, the convergence speed is relatively better than the conventional hill climbing algorithms. The reliability of the algorithms are tested for various homogeneous and heterogeneous irradiation conditions on a 3 × 3 array. The simulated results and the comparison between the two algorithms in terms of reduction ratio are presented. Keywords Photovoltaic array · Homogeneous irradiation · MPPT · Dichotomous search · Golden section search

1 Introduction Solar photovoltaic (PV)-based energy conversion systems use PV array to convert Sun’s light energy into electrical energy. The term conversion efficiency is used to quantize the amount of energy being converted. The conversion efficiency of a typical single junction PV cell is relatively low and it is material dependent. The widely used PV material is silicon and the efficiency depends on the crystalline structure. Polycrystalline Si solar cell has efficiency around 14–19% while the amorphous Si cell has an efficiency of about 6%. The overall efficiency of PV-based system depends not only on the efficiency of the PV array but also on the environmental factors and climatic conditions. The electrical output of the PV cell depends strongly on factors

R. Ramaprabha (B) · S. Malathy SSN College of Engineering, Rajiv Gandhi Salai, Kalavakam 603103, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_14

165

166

R. Ramaprabha and S. Malathy

like irradiation and temperature. The photocurrent generated by the PV cell falls with irradiation, and hence, there is a corresponding reduction in the output power [1]. The change in irradiation intensity intercepted by the PV panels is mainly due to the shading cast by clouds, trees or nearby structures. Improper maintenance of PV panels results in dust accumulation and it also reduces the interception of irradiation. In some circumstances, the irradiation intercepted by the entire PV array is uniform and its voltage–power (U–P) characteristic will exhibit single peak. The situation is common in case of smaller arrays. Also, in building applied PV systems, utmost care is taken in the planning stage itself and the panels are installed in such a way that they are free from the shadings that are cast by the nearby structures. Even if the panels are shaded by clouds, the shading will be homogeneous considering the size of the array. In such circumstances, the U–P characteristics will exhibit single maximum and conventional MPPT algorithms work well. However, in case of larger PV arrays, often, part of the array is shaded and the case is referred as partial shaded condition. Though the panels are installed in shade free areas, the panels or part of them are shaded by bird litters, nearby structures, passing clouds, etc. Under such circumstances, the U–P characteristic curve exhibits multiple peaks and the U–I characteristic curve exhibits multiple steps [2–4]. This is because of the fact that the shaded panels are excluded from the other serially connected panels by the bypass diodes. The situation is also common in urban residential installations where shading is unavoidable due to space limitations. The maximum power tracking [5–7] is difficult under these circumstances, as the conventional algorithms oscillate around the first peak they detect. This results in reduced power generation and underutilization of PV source. Many algorithms have been proposed in the literature to track the global peak amid the local peaks [8–10]. This chapter presents and compares two such global peak detecting algorithms that function well under both homogeneous and heterogeneous irradiation conditions. Both the algorithms are two stage algorithms, where the first stage reduces the search interval to the vicinity of the global peak and the second stage detects it accurately. The second stage of both the algorithms is based on ‘divide and discard’ strategy, where a significant portion of the search interval is eliminated at the end of each of the iteration [11]. The algorithms and their functionalities are explained in detail in the following sections. The performances of both the algorithms is compared for a 3 × 3 array and the results are presented.

2 Maximum Power Point Tracking Maximum power will be transferred from the PV source to the load when the corresponding impedances are matched. The impedance of the PV source varies with environmental conditions, and as a result, the impedances are to be matched dynamically to ensure maximum power extraction. The photocurrent of PV panel falls with irradiation and so does the power. Homogeneous irradiation conditions result in single peak. The bypass diodes are inactive as all the panels in the array intercept

Hybrid Algorithms to Track Peak Power in Solar PV Array …

a

b 400

PV Power (W)

PV Voltage (V)

60 40 20 0 200

167

400

600

800

Irradiation (W/sq.m)

1000

300 200 100 0 200

400

600

800

1000

Irradiation (W/sq.m)

Fig. 1 a Change in V pp with irradiation. b Change in Ppp with irradiation

similar irradiation. The variation in the maximum power (Ppp ) with irradiation and the voltage (V pp ) at which it occurs is depicted in Fig. 1. The voltage and current that correspond to the peak/global peak are represented as V pp and I pp , respectively. It can be seen from the figure that Ppp reduces significantly with the irradiation (from 335.3 W at 1000 W/m2 to 50.73 W at 200 W/m2 ). The variation in V pp usually lies between 70 and 85% of the open-circuit voltage of the PV array (49.8 V at 1000 W/m2 –44.4 V at 200 W/m2 ).To track the peak under homogeneous irradiation conditions, single stage tracking algorithms are sufficient. The tracking speed can be enhanced by properly fixing the search interval. Heterogeneous irradiation conditions result in multiple peaks in the U–P characteristic curve and the peak power depends on various factors including interconnection scheme, shade intensity and shade geometry. Heterogeneous irradiation conditions or partial shaded conditions are quite normal in PV-based energy conversion systems. Hence to track the global peak among the several other local peaks, hybrid tracking algorithms are required. Two such algorithms namely dichotomous search GMPPT (DSG) and golden section search GMPPT (GSSG) are considered in this chapter [12]. Both the algorithms are two staged algorithms, where the first stage reduces the search interval while the second stage accurately identifies it. The first stage of the algorithms should narrow down the search interval to the vicinity of the global peak. One way to do this is to sweep the entire search interval at significant regions. The maximum number of peaks in U–P curve of an array depends on the size of the array (number of panels in series string) and the number of bypass diodes per panel. If ‘m’ is the number of panels in string and one bypass diode is provided in each panel, then the search interval is fixed as (0, mV oc ) and the peaks are likely to occur at multiples of ‘kV oc_panel ,’ where ‘k’ lies between 0.7 and 0.85. The entire search interval is searched at these significant points (kV oc_panel , 2kV oc_panel , 3kV oc_panel , mkV oc_panel ) and the region of maximum power is identified. The area around the identified region forms the search interval for the second stage. This interval will have single peak and the second stage will accurately identify it; thus, the first stage of both the algorithms is based on fractional voltage strategy and second stage is based on ‘divide and discard’ strategy. The pseudocodes of both the algorithms are presented below.

168

R. Ramaprabha and S. Malathy

2.1 DS GMPPT The DSG algorithm identifies the peak of a unimodal function (function with single peak) in a given interval (a, b) by repeated division and removal. The two search points x 1 = c − ε and x 2 = c + ε are inserted on either side of the midpoint ‘c’ of the search interval. The resulting sections (a, x 1 ) and (x 2 , b) are shown in Fig. 2. The function is evaluated at the probe points x 1 and x 2 and one of the sections is eliminated. Section (a, x 1 ) is eliminated if f(x 1 ) < f(x 2 ) and the other section is eliminated otherwise. Thus, at the end of the iteration, the search interval is reduced. The search is continued till the interval shrinks to the global peak point. After tracking the peak, the algorithm measures the power periodically to identify change in irradiation conditions. New search is initiated if there is significant deviation in power. The algorithm is given below. Algorithm - DSG Input: Size of PV array (m×n), k,Voc-panel , search interval Step 1. Measure power Step 2. If ΔP >Pcritical, go to step 4 and start new tracking. Else go to step 1 Step 3. Initialize the search interval (0, mVoc-panel) Step 4. Set Vref_i= kVoc_panel and measure Power (Pi) Step 5. Apply voltage perturbation Step 6. Vref-i+1= Vref_i +kVoc_paneland measure Power (Pi+1) Step 7. Vref-max = Vref-i ;if Pi> P i+1elseVref-max = Vref-i +1 Step 8. Continue perturbation till 'i' is equal to 'm'; End of stage 1 Fix search interval around Vref-max. Step 9. Fix centre point c and probe points x1= c- ε , x2= c +ε Step 10. Measure the power P1 and P2 at the probe points Step 11. If P1> P2, shift left; Else shift right Step 12. If (P1 - P2 ) > Δ, Fix new search interval; repeat steps 9-12 Step 13. Else end of stage 2 Step 14. Wait for change in irradiation ; go to step 1

The progress in tracking the peak by the algorithm is presented in Fig. 3. Fig. 2 Search interval of DSG

Hybrid Algorithms to Track Peak Power in Solar PV Array …

169

Fig. 3 Progress in tracking

Fig. 4 Search interval of GSSG

Fig. 5 Progress in tracking by GSSG

a

52

50

Vref (V)

Vref (V)

51

49 48 47

b

50 48 46

0

0.2

0.4

0.6

Time (s) Fig. 6 Reference voltage a DSG. b GSSG

0.8

0

0.2

0.4

Time (s)

0.6

0.8

170

R. Ramaprabha and S. Malathy

a

b 300

300

Pmax

Vpv

Vout

200

Vout Vpv

100

100 0

Pout

200

0 0.4

0.2

0

0.6

0.8

0

0.2

0.4

0.6

0.8

Time (s)

Time (s) Fig. 7 Results a DSG. b GSSG 350

P1 = 335.3 W P2 = 155.8 W P3 = 104.2 W P4 = 299.5 W

PV Power (W)

300 250 200

P1 P4 P2

150 100

P3

50 0

0

10

20

30

40

50

60

70

PV Voltage (V) Fig. 8 U–P curve for different irradiation/shading conditions

2.2 GSS Gmppt The second stage of the second algorithm (GSSG) is based on golden section search [12]. This algorithm also works on unimodal functions. As the first stage of the algorithm narrows down the search interval around the vicinity of the global peak, the interval will have single peak. Two probe points x 1j = b − L j * and x 2j = a + L j * are inserted, where L j * = L o /γ j , L o is the initial search range and γ is the golden ratio. The search interval and probe points are shown in Fig. 4. The function is evaluated at both the search points and one section is eliminated at the end. The interval is reduced from (a, b) to (a, x 2 ) if f(x 1 ) > f(x 2 ) or to (x 1 , b) otherwise as shown in Fig. 5. Thus, the search interval is reduced in each of the iterations and new search points are inserted and the search continues till the interval shrinks to the global peak point. The algorithm is given below.

Hybrid Algorithms to Track Peak Power in Solar PV Array …

171

Algorithm - GSSG Input: Size of PV array (m×n), k,Voc-panel , search interval Step 1. Measure power Step 2. If ΔP >Pcritical, go to step 4 and start new tracking. Else go to step 1 Step 3. Initialize the search interval (0, mVoc-panel) Step 4. Set Vref_i= kVoc_panel and measure Power (Pi) Step 5. Apply voltage perturbation Step 6. Vref-i+1= Vref_i +kVoc_paneland measure Power (Pi+1) Step 7. Vref-max = Vref-i ;if Pi> P i+1elseVref-max = Vref-i +1 Step 8. Continue perturbation till 'i' is equal to 'm'; End of stage 1 Fix search interval around Vref-max. Step 9. Fix probe points x1j = b-Lj* and x2j = a+Lj* Step 10. Measure the power P1 and P2 at the probe points Step 11. If P1> P2, shift left; Else shift right Step 12. If (P1 - P2 ) > Δ, Fix new search interval; repeat steps 9-12 Step 13. Else end of stage 2 Step 14. Wait for change in irradiation ; go to step 1

3 Results and Discussions The performances of the two hybrid algorithms are compared under standard test conditions (STC) and shaded conditions. The shaded conditions include both homogeneous and heterogeneous shading conditions. The size of the array is small in case of residential building applied PV systems. These smaller arrays are usually placed on the rooftop and care is taken that the panels are not shaded by other nearby structures, cables or trees. Even if the panels are shaded by clouds, the shading will be homogeneous considering the size of the array. In such circumstances, the VP characteristics will exhibit single maximum and first stage of the algorithms is not required. It is enough to initiate the second stage alone. The performance of both the algorithms under STC and homogeneous shaded conditions is compared.

3.1 Homogeneous Shaded Conditions The 3 × 3 array considered in this chapter would generate a maximum power of 335.3 W at 49.8 V at standard irradiation condition of 1000 W/m2 . The reference voltages generated by both the DSG and GSSG algorithms are presented in Fig. 6. This reference voltage is compared with the actual PV array voltage and the error is fed as input to the controller. The controller adjusts the duty cycle of the gating pulses to the power converter so as to match the impedance between the source and the load.

172

R. Ramaprabha and S. Malathy

Power (W)

Fig. 9 Results of DSG

300

335.3 W

200

299.5 W

100 0

155.8 W 0

2

4

104.2W 6

8

PV voltage (V)

Time (s) 60 40

49.9 V

48.08 V 46.86 V

20 0

0

2

4

49.54 V

6

8

Time (s)

The variation in the PV voltage and the power that is extracted from the array by both the algorithms are presented in Fig. 7. Pmax represents the power extracted from the PV source, V out represents output voltage and V pv represents array voltage. The power extracted by both the algorithms is 335.3 W. The shading that is cast by the clouds can be considered homogeneous as the size of the array is very small in residential rooftop PV systems. The algorithm should be capable of tracking the change in the environmental (irradiation) conditions and operate the PV array at its optimal point. Both the DSG and GSSG algorithms are tested for changing irradiations conditions (homogeneous) and the results are compared. The irradiation levels are reduced from 1000 to 500 W/m2 and further to 355 W/m2 and finally increased to 900 W/m2 . The U-P curve of the 3 × 3 array for this homogeneous irradiation conditions is plotted in Fig. 8. The peak power reduces with the irradiation intensity. The U–P curves exhibit single peak as the irradiation is homogeneous, i.e., all the panels in the PV array receive same irradiation.Hence, it is enough if only the second stages of the DSG and GSSG algorithms are initiated. The reference voltages generated by both algorithms are presented in Figs. 9 and 10. Both the algorithms could track the change in irradiation conditions and the maximum power is extracted from the array as it is evident from the simulated results.

3.2 Heterogeneous Shaded Conditions The effectiveness of these two hybrid algorithms is tested under heterogeneous irradiation conditions also as it is most common in larger PV arrays. The arrays may be

Hybrid Algorithms to Track Peak Power in Solar PV Array …

Power (W)

Fig. 10 Results of GSSG

173

300 335.3 W

200

299.5 W

100 0

155.8 W

0

1

2

3

104.2W

4

5

6

7

PV voltage (V)

Time (s) 60 40

49.9 V

48.08 V 46.86 V

49.54 V

20 0

0

1

2

3

4

5

6

7

Time (s)

Fig. 11 Heterogeneous irradiation conditions

partially shaded by clouds, bird litters or accumulation of dust due to improper maintenance. If the array is partially shaded, the tracking algorithm should be capable of tracking the global peak amidst the local peaks. The 3 × 3 array considered in this work is partially shaded as shown in Fig. 11a and the corresponding U–P curve is presented in Fig. 11b. The array has two panels in its last row shaded at an intensity of 310 W/m2 . The first two non-shaded rows generate the rated current while the third row generates less. The V–P curve thus exhibits two peaks. The first stage of both the DS and GSS algorithm involves three searches to reduce the initial search interval to the vicinity of the global peak. The second stage of the algorithms locates the global peak accurately. The variation in the PV voltage in accordance with the reference voltage generated by the algorithms as the algorithms track the change in irradiation from partially shaded condition (Ppp of 232.7 W at V pp of 34.5 V) to standard conditions (Ppp of 335.3 W at V pp of 49.8 V) is presented in Figs. 12 and 13. Further, efficiency of the algorithm is quantified in terms of reduction ratio (RR) which gives the ratio by which the search interval is reduced in ‘n’ iterations. Lower

174

R. Ramaprabha and S. Malathy

Fig. 12 Results of DSS

Pmax

300

Vout

Vpv

200 100 0

3

2

1

0

4

Time (s)

Fig. 13 Results of GSSG

Pmax

300

Vout

200

Vpv

100 0

0

0.5

1

1.5

2

2.5

3

3.5

Time (s)

the ratio, better the tracking speed and efficiency. The reduction ratio is defined as RR =

Length of interval of uncertainity after n experiments (L n ) Length of intial inerval of uncertainity (L o )

(1)

The reduction ratios of all the two algorithms are tabulated in Table 1. The number of iterations required by all the two algorithms for a given error tolerance is calculated as error tolerance in% Ln = 2 100

(2)

The RR of GSS algorithm is (0.618n−1 ), where n is the number of iteration. With this RR, the GSS algorithm converges within ±5% of error tolerance after 6 iterations and to ±1% of error tolerance after 14 iterations. The DS algorithm with the RR of 1/2n/2 , required 6 and 12 iterations to converge to ±5% and ±1% of error Table 1 Reduction ratio of DSG and GSSG

Algorithm Initial interval Final interval GSS

L0

DS

L0

Reduction ratio

(0.618)n−1 L n =  L 0 (0.618)n−1   1 L n = L 0 1n n 22

22

Hybrid Algorithms to Track Peak Power in Solar PV Array …

175

tolerance respectively. These two algorithms require three (array has three rows) more iterations (first stage) to narrow down the search interval. This comparison is presented in Table 2. It can be inferred from the above results and tables that the DSG MPPT algorithm performs better in terms of accuracy and speed.

4 Conclusion Change in irradiation conditions is unavoidable as the power generation by the PV array is dependent on environmental conditions. This chapter has presented and compared two hybrid peak detecting algorithms. The algorithm and its functionality are explained in detail. The tracking ability of the two hybrid algorithms is validated for different shading conditions (standard, homogeneous and heterogeneous) through MATLAB-based simulations. Further, the reduction ratio is considered as a performance index to compare the tracking efficiency of the algorithms. It has been inferred that the DSG algorithm has better tracing efficiency and speed compared to the GSSG algorithm. For the considered case of 3 × 3 array, the DSG algorithm takes 12 iterations for homogeneous case and 15 iterations for heterogeneous irradiation conditions. The GSSG on the other hand requires 14 and 17 iterations respectively to track under homogeneous and heterogeneous conditions.

n

22

1

1 ≤ 0.1; n ≥ 6 n

22

GSS

DS

5% error

(0.618)n−1 ≤ 0.1; n ≥ 6

RR

(0.618)n−1

Algorithm

Table 2 Number of iterations of DSG and GSSG 1% error 1 ≤ 0.02; n ≥ 12 n

22

(0.618)n−1 ≤ 0.02; n ≥ 14 3

3

Iterations in first stage

Total iterations for 1% error 15

17

176 R. Ramaprabha and S. Malathy

Hybrid Algorithms to Track Peak Power in Solar PV Array …

177

References 1. Walker L, Hofer J, Schlueter A (2019) High-resolution, parametric BIPV and electrical systems modeling and design. Appl Energy 238:164–179 2. Lyden S, Haque ME (2019) Modelling, parameter estimation and assessment of partial shading conditions of photovoltaic modules. J Modern Power Syst Clean Energy 7(1):55–64 3. Tripathi AK, Aruna M, Murthy CS (2019) Performance of a PV panel under different shading strengths. Int J Ambient Energy 40(3):248–253 4. Malathy S, Ramaprabha R (2015) Comprehensive analysis on the role of array size and configuration on energy yield of photovoltaic systems under shaded conditions. Renew Sustain Energy Rev 49:672–679 5. Sikder PS, Pal N (2017) Incremental conductance based maximum power point tracking controller using different buck-boost converter for solar photovoltaic system. Rev Roum Sci Techn Électrotechn Et Énergy 62(3):269–275 6. Esram T, Chapman P (2007) Comparison of photovoltaic array maximum power point tracking techniques. IEEE Trans Energy Conv 22(2):439–449 7. Lyden S, Haque ME (2015) Maximum power point tracking techniques for photovoltaic systems: a comprehensive review and comparative analysis. Renew Sustain Energy Rev 52:1504–1518 8. Hashim N, Salam Z (2019) Critical evaluation of soft computing methods for maximum power point tracking algorithms of photovoltaic systems. Int J Power Electron Drive Syst 10(1):548– 561 9. Bahrami M, Gavagsaz-Ghoachani R, Zandi M, Phattanasak M, Maranzanaa G, NahidMobarakeh B, Pierfederici S, Meibody-Tabar F (2019) Hybrid maximum power point tracking algorithm with improved dynamic performance. Renew Energy 130:982–991 10. Al-Gizi A, Craciunescu A, Fadel MA, Louzazni M (2018) A new hybrid algorithm for photovoltaic maximum power point tracking under partial shading condition. Rev Roum Sci Tech Électrotech Et Énergy 63(1):52–57 11. Shao R, Wei R, Chang L (2014) A multi-stage MPPT algorithm for PV systems based on golden section search method. In: Twenty-ninth annual IEEE applied power electronics conference and exposition (APEC), pp 676–683 12. Malathy S, Ramaprabha R (2018) A two-stage tracking algorithm for PV systems subjected to partial shading conditions. Int J Renew Energy Res 8(4):2249–2256

Financial Analysis of Diesel and Solar Photovoltaic Water Pumping Systems M. Pandikumar and R. Ramaprabha

Abstract This chapter presents the financial analysis of the existing diesel-operated water pumping systems (WPS) used in the remote areas of Tamilnadu, India. The cost analysis of the conventional WPS (diesel pump) is compared with the solar photovoltaic water pumping system (SPV WPS). The SPV WPS uses either an induction motor (IM) or a brushless DC (BLDC) motor. The induction motor powered SPV WPS can be considered as an alternate solution for replacing the existing electric motor pumps and the diesel powered pumps. In addition, the comparison extends to the enhanced BLDC motor operated SPV WPS. The diesel-operated system and AC motors such as IM and BLDC motor operated are compared based on the life cycle cost of these systems. The cost analysis for the three WPS shows that BLDC operated SPV WPS is a cost-effective solution for the induction motor operated SPV WPS and the diesel powered WPS. Keywords Solar water pump · Diesel pump · Financial analysis · Life cycle cost

1 Introduction The technology is growing at a faster rate to ease the needs of the mankind. But the main focus of the people is the development of the agricultural sector. Renewable energy sources are being introduced to overcome the present technological shortage and power shortage in the agricultural sector. Of the main focus, providing water to the agricultural fields through water pumps for the entire period of the crop growing period from the deep wells is considered a major factor [1, 2]. Most of the existing WPS in rural India used for domestic and agricultural needs is diesel powered. The remaining WPS are electrically powered, mostly with induction motors. The drawbacks of the existing machines are overcome with the advancement in the development of special electrical machines such as the BLDC. These special electrical M. Pandikumar (B) · R. Ramaprabha Sri Sivasubramaniya Nadar College of Engineering, Rajiv Gandhi Salai, Kalavakkam, Chennai, Tamil Nadu 603110, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_15

179

180

M. Pandikumar and R. Ramaprabha

Fig. 1 Block diagram of the conventional diesel WPS

motors have many advantages over the conventional induction motors and may be considered for use with solar powered WPSs (WPSs). The main objective of this chapter is to analyze a conventional diesel WPS and a solar powered AC WPS and compare them based on the life cycle cost (LCC) [3]. The chapter presents the comparative analysis of a diesel WPS, a solar photovoltaic IM WPS and a solar photovoltaic BLDC WPS.

2 Conventional and Solar Powered WPS Most of the WPS used in the remote areas or areas with frequent power outages use diesel or kerosene operated WPS to feed water for domestic usage and for agriculture usage [4]. Figures 1 and 2 represent the block diagram of the conventional and solar powered WPSs. One of the effective ways of SPV WPS is to connect the load directly to the PV system [5]. With a direct connection between the PV array and the pump, water can be pumped directly or stored in an elevated storage tank during the sunlight hours. The stored water flows to the agriculture field with the help of gravity. The overall efficiency of WPS setup varies and depends on factors like the environment, controllers, type of motor and other components [6]. Even though the investment cost is high, the investment can be got back within a stipulated period of 5–10 years.

3 Parameters Considered for the Financial Analysis Most of the crops cultivated in India use 3000–20,000 m3 /ha of water. Table 1 shows

Financial Analysis of Diesel and Solar Photovoltaic Water …

181

Fig. 2 Block diagram of the solar powered WPS

Table 1 Water requirement of seasonal crops Crop

Average growing period (days)

Average amount of water needed for the growing period (mm)

Average value of water (m3 /ha)

Rice

120

575

5600

Wheat

135

550

4550

Maize grain

150

650

3650

Onion

180

450

4250

Potato

125

600

4000

the water requirement of the various seasonal crops in India [7, 8]. In this work, the water requirement for rice cultivation in an area of 1 ha (2.47 acres) for 120 days is considered. Hence, the water required for an average of 120 days is 5,600 m3 /ha/day [9]. Water required for the complete growing period = 5,600 × 1,000 L/ha per 120 days. Daily flow rate = 5,600 × 1,000/120 = 46,667 L/ha/day. Flow rate per minute = 46,667/(5.3 × 60) = 146 L/ha/min. Flow rate per second = 146/60 = 2.44 L/ha/s. The total dynamic head (TDH) of the solar powered WPS and the diesel powered WPS is considered as 25 m. The pump is selected considering the induction and BLDC motor pumps to deliver the water. Table 2 gives the rating selected for operating the solar pump that delivers 47,000 L/day. The total power delivered by the PV array is (300 × 4.3) = 1290 Wp for the SPV WPS [10–12]. Table 3 shows the parameters used for the cost analysis of the conventional and solar powered WPS.

182

M. Pandikumar and R. Ramaprabha

Table 2 Selection of power rating of the pump Type of motor

Phase (Ph)

Power (HP)

Voltage (V)

Current (A)

Induction motor

3

1

300

5

BLDC motor

3

1

300

4.3

Diesel pump

Table 3 Main input parameters for the cost analysis

1

S. No

Parameters

Value

1

Pumping head

25 m

2

Daily pumping rate

46.46 m3 /day

3

Daily average solar irradiance level

5.3 kWh/m2 /day

4

Tracking angle of the solar panels

13.5°

5

Type of motor

Induction motor/BLDC motor

6

Diesel pump selection

Long / short lifespan

7

Pumping hours/day for diesel pump

5

8

Cost of fuel

Rs. 50/L

9

Annual diesel price escalation

4%

10

Project life

20 years

For the practical considerations of the proposed work, the solar panels taken into consideration are from SOLKAR which has power specification of 37.08 W with open-circuit voltage & short-circuited current are 21.24 V and 2.55 A. The voltage and current of the solar panel at maximum power conditions are of 16.56 V and 2.25 A. The system size for the SPV WPS and the diesel-operated WPS is taken to be 1HP. Since the power delivered by the solar panel is of 37 W, a total of 20 number of solar panels are connected in series to form an array to deliver the required power. Two sets of series connected panels are connected parallel to each other to meet the current requirements of the SPV WPS. The diesel WPS is of 1HP power rating delivers 160 L/min from the head level of 5–30 m.

4 Life Cycle Cost (LCC) of the Conventional and SPV WPS Figure 3 shows the LCC of the system calculated over a 25-year period [13, 14]. These costs include:

Financial Analysis of Diesel and Solar Photovoltaic Water …

183

Fig. 3 Overview of the LCC for the SPV and the diesel water pumping systems

• The initial upfront cost; • The operating costs (operational cost, inspection cost of solar water pumps and fuel cost for the diesel pumps); • Maintenance costs; and • Replacement costs. The LCC analysis method compares the economic costs for the SPV and the diesel pumping systems. These costs include the initial investment cost and the additional costs, which include cost for the operation and maintenance and the replacement. The calculations are carried out without considering the government subsidy and other benefits of solar power installations. The non-subsidy consideration makes the capital cost of the system higher. For the diesel-operated system, the fuel consumed per annum is calculated by the duration of operation of the pump. The maintenance and replacement costs are common to the SPV WPS and dieseloperated pump. The maintenance schedule and details are dependent on the technology employed. The service interval depends on the pump system used, water quality and depth of installation. On considering the replacement cost over a period of operation of motor and pump will be the same as that of the initial cost. Table 4 shows the initial costs for the induction motor controlled SPV WPS and BLDC motor controlled SPV WPS [15–24]. Figure 4 shows the LCC of the SPV WPS controlled by the induction motor

184

M. Pandikumar and R. Ramaprabha

Table 4 Initial costs for the IM SPV WPS and BLDC SPV WPS PVP system components

Induction motor controlled SPV WPS

BLDC controlled SPV WPS

PV array power for 25 m TDH for delivering 46.67 m3 /day

2050 Wpeak

845 Wpeak

PV array

Rs. 85,565.00

Rs. 34,024.00

PV array structure

Rs. 4,899.00

Rs. 2,430.00

PV pump with controller

Rs. 36,000.00

Rs. 68,500.00

Pipe, cable and rope

Rs. 3,438.00

Rs. 2,993.00

Accessories

Rs. 10,392.00

Rs. 8,636.00

Installation

Rs. 5,199.00

Rs. 2,730.00

Total PVP installation cost (value added tax (VAT) included)

Rs. 1,67,318.00

Rs. 1,37,210.00

300000.0

Cost in Rupees

250000.0 200000.0 150000.0 100000.0

IM (Repl:7 yrs, Main: 2.5 yrs) IM (Repl:8 yrs, Main: 4 yrs)

50000.0

IM (Repl:10 yrs, Main: 5 yrs) 0.0 1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20

Years Fig. 4 LCC of the induction motor operating at 25 m head and delivering 46 m3 /day

operating at 25 m head and delivering 46 m3 /day with different replacement periods and variable maintenance schedule. Figure 5 shows the LCC of the BLDC motor operating at 25 m head and delivering 46 m3 /day with different replacement periods and variable maintenance schedule. The LCC analysis of the induction powered SPV WPS operating at 25 m head with ten years of replacement period and five years of maintenance cycle gives the least LCC of the three machines. The LCC analysis of the BLDC powered SPV WPS as shown in Fig. 5 operating at 25 m head with 18 years of replacement period and ten years of maintenance cycle

Financial Analysis of Diesel and Solar Photovoltaic Water …

185

250000.0

Cost in Rupees

200000.0 150000.0 100000.0 BLDC (Repl: 10 yrs, Main: 6 yrs) BLDC (Repl: 15 yrs, Main: 10 yrs)

50000.0

BLDC (Repl: 18 yrs, Main: 10 yrs) 0.0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20

Years Fig. 5 LCC of the BLDC motor operating at 25 m head and delivering 46 m3 /day

gives the least LCC. The BLDC powered SPV WPS with high period of operation and maintenance shows a better option for SPV WPS. The maintenance cost of the SPV pumping systems is applicable to both the PVP and diesel pumps. The maintenance schedule and service interval depends on the pump systems used. In the case of the induction motor operated SPV WPS, the service interval is five years from the installation period as given in Table 5. whereas in the case of BLDC operated SPV WPS is ten years as given in Table 6. Table 5 Cost breakdown of the LCC of IM operated SPV WPS Year

0

Initial cost—PV array of 2050 W peak

1,67,318

Operating cost

5

10

15

20

1,67,318 174.72 per year

Replacement of main components

3,494 9,600

Maintenance and service

Total

4,178

9,997 9,578

49,596 23,756

Table 6 Cost breakdown of the LCC of BLDC operated SPV WPS Year

0

Initial cost—PV array of 845 W peak

1,37,210

Operating cost

5

10

20

Total 1,37,210

174.72 per year

3,494 39,680

Replacement of main components Maintenance and service

18

5,438

39,680 3,673

23,756

186

M. Pandikumar and R. Ramaprabha

Table 7 LCC of diesel pump, IM SPV WPS and BLDC SPV WPS Initial cost (Rs.) Diesel WPS

Operating cost (Rs.)

44,104

Maintenance cost (Rs.)

Replacement cost (Rs.)

Total (Rs.)

2,17,065

1,38,273

78,116

4,77,558

IM SPV WPS 1,67,318

3,494

23,756

49,596

2,44,165

BLDC SPV WPS

3,494

9,112

39,680

18,949

1,37,210

600000.0

Diesel WPS IM SWPS

LCC cost in rupees

500000.0

BLDC SWPS

400000.0 300000.0 200000.0 100000.0 0.0 0

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20

Years Fig. 6 LCC of the diesel WPS, IM SPV WPS and BLDC SPV WPS operating at 25 m TDH

The LCC cost of the induction motor and the BLDC motor operated solar water pump can be compared with the diesel pump operating at 25 m TDH and delivering 46.67 m3 /day. Table 7 gives the calculated values of the LCC for the diesel and IM/BLDC SPV WPS. The LCC analysis of the IM & BLDC operated SPV WPS and diesel powered WPS is shown in Fig. 6. The LCC analysis of the BLDC powered SPV WPS gives a minimum of 4.2 years payback period when compared with the diesel powered WPS. The LCC analysis of three systems, BLDC powered SPV WPS has a higher payback when compared to IM powered SPV WPSs and diesel WPS.

5 Conclusion This work shows that BLDC operated SPV WPS delivers 46.67 m3 /day of water at 25 m TDH more efficiently than other WPS. The results are compared with the induction motor operated SPV WPS and the results show that the overall cost for

Financial Analysis of Diesel and Solar Photovoltaic Water …

187

this system is higher than the BLDC operated SPV WPS. The LCC cost analysis of the BLDC operated SPV WPS is lower than the induction motor powered SPV WPS for the period of 20 years. The BLDC operated SPV WPS when compared with the diesel-operated pump has an LCC crossover at 4.2 years of operation while the induction motor system crossover is at 5.6 years of operation with the diesel-operated pump. This clearly points out that BLDC motor operated SPV WPS is better when compared to the induction motor operated SPV WPS and diesel-operated WPS for the operating period of 20 years.

References 1. Maurya VN, Ogubazghi G, Misra BP, Maurya AK, Arora DK (2015) Scope and review of photovoltaic solar water pumping system as a sustainable solution enhancing water use efficiency in irrigation. Am J Biol Environ Stat 1(1):1–8 2. Khlifi MA (2015) Study and control of photovoltaic water pumping system. J Electric Eng Technol 11(1):709–718 3. Girma M, Assefa A, Molinas M (2015) Feasibility study of a solar photovoltaic water pumping system for rural Ethiopia. AIMS Environ Sci 2:697–717 4. Jenkins T (2014)Designing solar water pumping systems for livestock, cooperative extension service—engineering. New Mexico Res Netw 1–12 5. Elgendy MA, Zahawi B, Atkinson DJ (2010) Comparison of directly connected and constant voltage controlled photovoltaic pumping systems. IEEE Trans Sustain Energy 1(3):184–192 6. Raghav MB, Bhavya KN, Suchitra Y, Rao GS (2013) Design of solar power based water pumping system. Int J Eng Res Technol (IJERT) 2:333–339 7. Doorenbos J, Pruitt WO (1977) Crop water requirements, Food and Agriculture Organization of the United Nations revised report 8. NSW Farmers, GSES (2015) Solar-powered pumping in agriculture: a guide to system selection and design 9. Chikuni E (2012) Program-assisted sizing of a photovoltaic-powered water pumping system’. J Energy Southern Africa 23(1):32–38 10. Acakpovi A, Xavier FF, Awuah-Baffour R (2012) Analytical method of sizing photovoltaic water pumping system. In: IEEE 4th international conference on adaptive science & technology (ICAST). IEEE, pp 65–69 11. Abu-Aligah M (2011a) Design of photovoltaic water pumping system and compare it with diesel powered pump. Jordan J Mech Ind Eng (JJMIE) 5(3):273–280 12. Morales TD, Busch J (2010) Design of small photovoltaic (PV) solar-powered water pump systems. Technical Note No. 28, United States Department of Agriculture, Washington DC, pp 55–71 13. Alireza R, Asghar GS (2013) Technical and financial analysis of photovoltaic water pumping system for GORGAN, IRAN. Int J Cybern Inf (IJCI) 2(2):21–31 14. Kabade A, Rajoriya A, Chaubey UC (2013) Solar pump application in rural water supply—a case study from Ethiopia. Int J Energy Eng (IJEE) 3:176–182 15. Pandikumar M, Ramaprabha R, Muthu R (2015) Analysis of controllers for photovoltaic fed brushless DC motor based water pumping system. Appl Mech Mater 787 (Trans Tech Publications) 16. Pandikumar M, Ramaprabha R, Muthu R (2013) Design and modeling of photovoltaic system fed brushless DC motor. In: Advanced materials research, vol 768. Trans Tech Publications, pp 136–142

188

M. Pandikumar and R. Ramaprabha

17. Pandikumar M, Ramaprabha R, Muthu R (2015) Analysis of controllers for photovoltaic fed brushless DC motor based water pumping system. In: Applied mechanics and materials, vol 787. Trans Tech Publications, pp 838–842 18. Sontake VC, Kalamkar VR (2016) Solar photovoltaic water pumping system—a comprehensive review. Renew Sustain Energy Rev 59:1038–1067 19. Al-Waeli AH, El-Din MM, Al-Kabi AH, Al-Mamari A, Kazem HA, Chaichan MT (2017) Optimum design and evaluation of solar water pumping system for rural areas. Int J Renew Energy Res 7(1):12–20 20. Tiwari AK, Kalamkar VR (2018) Effects of total head and solar radiation on the performance of solar water pumping system. Renew Energy 118:919–927 21. Aliyu M, Hassan G, Said SA, Siddiqui MU, Alawami AT, Elamin IM (2018) A review of solar-powered water pumping systems. Renew Sustain Energy Rev 87:61–76 22. Al-Badi A, Yousef H, Al Mahmoudi T, Al-Shammaki M, Al-Abri A, Al-Hinai A (2018) Sizing and modelling of photovoltaic water pumping system. Int J Sustain Energ 37(5):415–427 23. Ali B (2018) Comparative assessment of the feasibility for solar irrigation pumps in Sudan. Renew Sustain Energy Rev 81:413–420 24. Raghuwanshi SS, Khare V (2018) Sizing and modelling of stand-alone photovoltaic water pumping system for irrigation. Energy Environ 29(4):473–491

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode Controller in PMSG-Based Wind Turbines M. Rajvikram and J. Vishnupriyan

Abstract The work reported in this research chapter is to implement the power maximization control technique toward the rectifier side of the PMSG-based gridconnected WECS. To prove that, proposed tuned two input fuzzy-based maximum power point tracking controller is better than the normal fuzzy-based single input MPPT controller. The simulation results of both the controllers are realized and the substantiating improvement in terms of real power produced across the grid is analyzed. It is found that the performance of tuned FLC is better. In most of the existing literatures, commonly occurring asymmetrical faults are taken for consideration to assess the transient study of the WECS. As a second part of analysis, in the proposed work, rarely occurring faults such as three-phase symmetrical fault (LLL) are also taken into account along with asymmetrical fault and the study is thoroughly conducted by injecting the fault towards the grid side in MATLAB Simulink environment. The motive of this study is that, though the symmetrical fault is rarely occurring, the negative impact it creates on the system is very high. It is a main factor that has to be considered while designing the controllers. SMC is proposed in order to bring the system back to the normal state during the period of fault occurrence for any kind of fault towards the grid side. This chapter throws light on a fact that even during grid fault conditions, the proposed tuned FLC and SMC holds good in coordination of system functions. The entire wind system is modeled in such a way that there is perfect coordination between tuned FLC and SMC during steady-state and transient-state conditions. Keywords Sliding mode proportional integral (PI) controller · Grid fault conditions · Tuned fuzzy logic controller (FLC) · Maximum power point tracking (MPPT) · Wind energy conversion system (WECS) · Permanent magnet synchronous generator (PMSG)

M. Rajvikram (B) Sri Venkateswara College of Engineering, Chennai, India e-mail: [email protected] J. Vishnupriyan Chennai Institute of Technology, Chennai, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_16

189

190

M. Rajvikram and J. Vishnupriyan

1 Introduction—Literature Surveys 1.1 A Subsection Sample Green energy sources are getting promoted drastically due to the low availability of tar and coal [1–4]. Researchers are more focused on renewable power production, especially working on how to improve the existing efficiency of the renewable power plants such as wind and photovoltaic [5–11]. The surveys are taken pertaining to the complete operation of a wind system during normal operating conditions and grid fault conditions. Some of the surveys are as follows, a novel method which has explained about both the design and analysis of power distribution to the grid [12]. Thorough evaluation has been made pertaining to the predictive model control and duty cycle modulation ratio for the generator side control of a grid-connected wind system. The field programmable gate array-based programmable controller has been designed in order to improve the real power producing capability of the grid-connected wind system. There are several techniques implemented which have eliminated the need for a direct control of nonlinear system [13]. Superconducting fault current limiter (SFCL) plays a vital role in suppressing the DC-link voltage oscillations when the grid-connected PMSG-based wind system is subjected to fault [14]. The generator used in the modeling of grid-connected WECS is squirrel cage induction generator (SCIG). The mathematical modeling of the controller design is done and the equations to obtain the maximum power at various wind speeds are formulated [15]. Grid voltage measurement methods have been proposed for control of reactive power. The author has brought out the stability analysis methods with respect to the control of real power and reactive power. The importance of PF and QV control loop is clearly explained [16]. Sliding mode control (SMC) is introduced to a variable structure system [17]. SMC is widely used for robustness toward variable input/output and parameter uncertainties. Power converters are naturally variable structure systems due to its switching function. Dynamic response of the SMC was investigated for DC–DC converters in [18]. Minimized total harmonic distortion in current and voltage waveform with constant switching frequency for reduced losses were carried out by the SMC [19]. Such properties promote the usage of SMC in commercial and industrial utilities through grid-connected converter [20]. In [21], the authors have investigated SMC in Z-source inverter. A method to diminish chattering phenomenon is to use hysteresis band in the forced switching pulse width modulator-based SMC, alongside there are many researches that are happening with respect to various inverter level designs, technology implementation in energy management and Indian perspectives of global renewable energy scenarios which are discussed in [22–27]. So far, none of the literature has addressed the investigation on SMC in PMSG-based wind turbine for the purpose of DC-link voltage control. This chapter presents the detailed analysis and demonstrates the dynamics of the SMC for wind energy conversion system integrated with the grid.

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

191

Fig. 1 Block diagram of PMSG-based grid-connected wind energy conversion system

2 PMSG-Based Wind Energy Conversion System Figure 1 represents the block diagram representation of PMSG-based grid-connected WECS. From Fig. 1, it can be inferred that RSC is experimented with FLC. The GSC is executed with SMC during the steady and transient operating conditions, respectively.

3 Analysis of Fault and Impact of Faults on Wind System 3.1 Methodology This section will clearly demonstrate the controller design concept. Here, the controller implemented toward the rectifier side is MPPT controller. MPPT controller method is implemented by means of tuned fuzzy logic controller, and during grid fault condition in order to protect the system from uncertainty conditions, the controller implemented toward grid side is sliding mode PI controller.

192

M. Rajvikram and J. Vishnupriyan

3.2 FLC Design FLC design is implemented toward the rectifier side during normal operating conditions of the grid to set the generator at an optimal speed that is the fuzzy logic controller output generates the speed reference by means of change in mechanical power and change in the wind speed of the turbine as inputs. The controller is modeled to extract the uttermost power during various wind speeds. Table 1 will give the division of various regions for both power and speed change. The conceptual circuit of the controller implementation toward rectifier side is shown in Fig. 2. Here, fuzzy logic rule bases tuning technique is used. Table 1 Fuzzy inputs and outputs  ωr/Pm

NL

NM

NS

Z

PS

PM

PL

NL

PL

PL

PM

NS

NM

NL

NL

NM

L

PM

PS

NS

NS

NM

NL

NS

PM

PS

PS

Z

NS

NS

PL

Z

NL

NM

NS

Z

PS

PM

PL

PS

NM

NS

NS

Z

PS

PS

PM

PM

NL

NM

NS

PS

PS

PM

PL

PL

NL

NL

NM

PS

PM

PL

PL

NM

PL

PM

PS

NS

NS

NM

NL

Fig. 2 Rectifier side torque control conceptual circuit with fuzzy logic controller

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

193

Number of Inputs: 2; Number of Output: 1; Input 1: Change in wind turbine speed; Input 2: Change in mechanical power; Output: Speed Reference of the generator.

4 Grid Side Control of a WECS—Fault Operating Condition Faults namely symmetrical and unsymmetrical faults are injected toward the grid side and the performance of the system is analyzed. SMC is implemented to protect the machine during fault operating conditions. The controller design is governed by the following set of equations. In SMC, the DC-link voltage tracing equation is taken into account, the value of nominal DC-link voltage is taken in closed loop and it was fed in closed loop with respect to the interconnected grid. During transient condition, SMC generates controlled direct axis reference current pulses to the inverter side. By doing so, the shoot up in the DC-link voltage during abnormal conditions can be minimized. DC-link voltage tracing error equation is given by ∗ ev = VDC − VDC

(1)

1 ∗ = B V i d − i s + dv − VDC c

(2)

where BV =

1 3 vd C 2 v DC ∗

(3)

With VDC ∗ is the DC-link voltage reference. BV Indicates the difference between the actual voltage and the reference voltage ev = u v + dv

(4)

where u v is the latest control input and modelled as follows, 1 u v = Bv i d − i s + dv + VDC ∗ c

(5)

Second SMC strategy is applied for the DC-link voltage reference such as u v = −k1 |ev |1/2 sgn(ev ) + w

(6)

w˙ = −k2 sgn(ev )

(7)

194

M. Rajvikram and J. Vishnupriyan

The command id is carried out from (5) and (6) such as id =

  1 1 −k1 |ev |1/2 sgn(ev ) + w + i s + VDC ∗ Bv c w˙ = −k2 sgn(ev )

(8) (9)

The analysis of the research work is subdivided into two parts. The first analysis is the modification of control design toward the rectifier side, and to extract the uttermost power at various wind speed, tuned fuzzy logic control technique is proposed. The second analysis is to give different type of faults toward the grid side, and to bring the system to the normal state from the abnormal state, SMC is proposed.

5 Results and Discussions This portion will clearly elucidate the simulation results pertaining to FLC and SMC implemented in PMSG-based WECS during steady-state and transient-state conditions.

5.1 Power Maximization Control—Steady-State Analysis The simulation outputs and waveforms obtained through Fig. 3 during the speed of wind at 12 m/s are shown. Table 2 shows that the power generation is high during higher the level of wind speed after the implementation of tuned fuzzy-based MPPT

Fig. 3 Grid-connected WECS simulation diagram

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

195

Table 2 Results of proposed model PMSG-based WECS Speed of wind (m/s)

Machine without controller (power in kW)

With fuzzy (one input) controller (power in kW)

With tuned fuzzy MPPT (two input) controller (power in kW)

6

10

80

90

9

15

90

100

12

17

110

140

15

19

140

170

18

21

170

200

controller compared to the fuzzy one input controller and the significant results obtained by means of the proposed controller during the speed of wind at 12, 15 and 18 m/s. Figures 4 and 5 represent the synchronized step-up transformer output voltage and current waveform at the wind speed of 12 m/s. The graph shown in Fig. 6 shows the power produced by the PMSG without MPPT controller, with fuzzy one input MPPT controller and with tuned fuzzy MPPT controller for each wind turbine speed, and the variation is clearly denoted.

2

10

4

1.5

Voltage (V)

1 0.5 0 -0.5 -1 -1.5 -2 0.5

0.51

0.52

0.53

0.54

0.55

Time (s)

Fig. 4 Step-up transformer output voltage waveform

0.56

0.57

0.58

0.59

0.6

196

M. Rajvikram and J. Vishnupriyan 5 4 3

Current (A)

2 1 0 -1 -2 -3 -4 -5 0.75

0.8

0.9

0.85

0.95

1

Time (s)

Fig. 5 Step-up transformer output current waveform

Fig. 6 Comparison graph showing the real power generated to the grid

5.2 SMC Implementation—Transient Conditions This section contains the simulation outputs that will clearly describe the performance of the system during grid fault conditions. Various faults namely symmetrical fault and unsymmetrical faults are given toward the grid side and the performance is analyzed during the period of fault duration. When the sliding mode PI controller is implemented toward the grid side control system, it improved the overall fault clearing capability of the system. Figure 7 shows the overall simulation diagram. It indicates the changes that are caused due to the fault injection toward the grid side and how the system is made stable with the implementation of the proposed schemes

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

197

Fig. 7 Grid-connected WECS simulation diagram with fault injected towards the grid side

Fig. 8 Subsystem of VDC regulator with sliding mode PI controller

can be inferred in this section. Figure 8 represents the subsystem of voltage regulator where the sliding mode PI controller is implemented.

5.3 Simulation Results—Fault Analysis with Sliding Mode PI Controller Implemented Towards the Grid Side Control System This section will give the simulation results of the symmetrical fault, namely LLL fault that is injected toward the grid side and the effectiveness of SMC used inside the V DC regulator in bringing the system from fault state to the normal state can be realized. Two different fault durations are given starting from half cycle to the maximum of five cycles of fault and the changes in the DC-link voltage across the capacitor, inverter side current and voltage waveforms are realized. In this section, fault duration of half cycle and five cycles is given toward the grid side and the performance of the system during different fault condition is simulated and the results

198

M. Rajvikram and J. Vishnupriyan 10

2

4

1.5

Voltage (V)

1 0.5 0 -0.5 -1 -1.5 -2

1

1.02

1.04

1.06

1.08

1.1

1.12

1.14

1.16

1.2

1.18

Time (s) Fig. 9 Subsystem of VDC regulator with sliding mode PI controller

are realized with the implementation of sliding mode PI controller. Figure 9 indicates the grid side voltage waveform; it can be observed that during LLL symmetrical fault, there is disturbance during the time period of 1.1–1.11 s; after the period of fault occurrence, the voltage level is brought back to the normal state. Figure 10a indicates the grid side current waveform; it can be observed that during LLL symmetrical fault, there is disturbance during the time period of 1.1–1.11 s; after the period of fault occurrence, the shoot up of current is about 2500 A, and after the period of fault occurrence, it is regulated and brought back to the normal value. Figure 10b indicates that there occurs a peak shoot up of about 4500 V during the period of fault occurrence 1.1–1.11 s. After the period of fault occurrence, the DC-link voltage is well regulated with the aid of sliding mode PI controller. 3000

5000 4500 4000 3500

1000

Vdc (V)

Current (A)

2000

0 -1000

3000 2500 2000 1500 1000

-2000

500 -3000

1 1.02 1.04 1.06 1.08 1.1 1.12 1.14 1.16 1.18 1.2 Time (s)

(a)

0

0

0.5

1

1.5 Time (s)

2

2.5

3

(b)

Fig. 10 a Subsystem of VDC regulator with sliding mode PI controller. b Subsystem of VDC regulator with sliding mode PI controller

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

199

Table 3 DC-Link voltage regulation with normal PI controller and sliding mode PI controller after the fault clearance Nominal DC-link voltage = 1150 V Type of fault

V DC after the fault clearance of half cycle (V)

V DC after the fault clearance of half cycle (V)

LLL

1139

1145

LLLG

1140

1146

LG

1140

1150

LLG

1140

1150

LL

1140

1150

5.4 Steady-State Value of DC-Link Voltage with Sliding Mode PI Controller After the Fault Has Been Removed See Table 3.

6 Conclusions In the first analysis, real power produced at the particular wind speed and the improved efficiency of the proposed tuned fuzzy logic controller is brought about clearly by means of the conceptual graph as shown in Fig. 6 with the tuned fuzzy-based MPPT Controller, with the fuzzy one input MPPT controller and without MPPT controller. From Table 2, it is concluded that the suggested controller-tuned FLC MPPT controller used in the rectifier control design during normal operating conditions of the grid generates higher power with the implementation of the proposed MPPT controller as the wind turbine speed increases and it is the best for high wind speed application. The results obtained are fairly encouraging and its proved to be more efficient. In the second analysis, fault is injected toward the grid side and the performance of the wind system is thoroughly studied during grid fault conditions. From Table 3, it can be summarized that the developed SMC is proved to be the best in terms of performance as it is bringing the value of DC-link voltage close to the nominal after the fault has been cleared compared to the normal PI controller used in the grid side control system.

References 1. Rajvikram M (2019) The motivation for renewable energy and its comparison with other energy sources: a review. Eur J Sustain Develop Res 3(1)

200

M. Rajvikram and J. Vishnupriyan

2. Rajvikram M, Renuga P, Swathisriranjani M (2016) Fuzzy based MPPT controller’s role in extraction of maximum power in wind energy conversion system. In: International conference on control, instrumentation, communication and computational technologies 3. Pandiyarajan A, Rajvikram M (2016) Electric arc furnace voltage flicker alleviation by unified power quality conditioner using PSCAD/EMTDC. In: International conference on control, instrumentation, communication and computational technologies 4. Vishnupriyan J, Manoharan PS (2018) Prospects of hybrid photovoltaic–diesel standalone system for six different climate locations in Indian state of Tamil Nadu. J Clean Prod 185:309– 321 5. Krishnamoorthy R et al (2020) An assessment of onshore and offshore wind energy potential in India using moth flame optimization. Energies 13(12):3063 6. Rajvikram M (2019) Role of dual input fuzzy controller for the better production of real power in wind system. J Electric Eng 19(2):44 7. Rajvikram M, Ghosh A, Mallick TK, Apoorva K, Meenal S (2019) Investigations on performance enhancement measures of the bidirectional converter in PV–wind interconnected microgrid system. Energies 12(14):672 8. Rajvikram M (2018) Solutions for voltage SAG in a doubly fed induction generator based wind turbine: a review. Power Res 14(1) 9. Rajvikram M, Leoponraj S (2018) A method to attain power optimality and efficiency in solar panel. Beni-Suef Univ J Basic Appl Sci 7(4):705–708 10. Rajvikram M, Leoponraj S, Ramkumar S, Akshaya H, Dheeraj A (2019) Experimental investigation on the abasement of operating temperature in solar photovoltaic panel using PCM and aluminium. Sol Energy 188:327–338 11. Rajvikram M, Sivasankar G (2019) Experimental study conducted for the identification of best heat absorption and dissipation methodology in solar photovoltaic panel. Sol Energy 193:283–292 12. Vishnupriyan J, Manoharan PS (2017) Demand side management approach to rural electrification of different climate zones in Indian state of Tamil Nadu. Energy 138:799–815 13. Zhang Z, Rodrígue J, Kennel R (2017) Advanced control strategies for direct-drive PMSG wind turbine systems: direct predictive torque control approaches. CPSS Trans Power Electron Appl 2(3) 14. Rajvikram M, Renuga P, Aravind Kumar G, Bavithra K (2016) Fault ride-through capability of permanent magnet synchronous generator-based wind energy conversion system. Power Res 12(3) 15. Mesemanolis A, Mademlis C, Kioskeridis I (2012) High-efficiency control for a wind energy conversion system with induction generator. IEEE Trans Energy Conv 27(4) 16. Yoo C-H, Chung Y, Yoo H-J, Hong S-S (2014) A grid voltage measurement method for wind power systems during grid fault conditions. Energies 7(11) 17. Tan SC, Lai YM, Cheung MK, Tse CK (2005) On the practical design of a sliding mode voltage controlled buck converter. IEEE Trans Power Electron 20(2):425–437 18. Tan SC, Lai YM, Tse CK, Cheung MK (2005) A fixed-frequency pulse width modulation based quasi-sliding-mode controller for buck converters. IEEE Trans Power Electron 20(6):1379– 1392 19. Guzman R, de Vicuña LG, Morales J, Castilla M, Matas J (2016) Sliding-mode control for a three-phase unity power factor rectifier operating at fixed switching frequency. IEEE Trans Power Electron 31(1):758–769 20. Pires VF, Martins JF, Hao C (2012) Dual-inverter for grid-connected photovoltaic system: modeling and sliding mode control. Sol Energy 86(7):2106–2115 21. Kumar N, Saha TK, Dey J (2017) Sliding mode control, implementation and performance analysis of standalone PV fed dual inverter. Solar Energy 155:1178–1187 22. Tan SC, Lai YM, Tse CK (2008) General design issues of sliding-mode controllers in DC–DC converters. IEEE Trans Ind Electron 55(3):1160–1174 23. Gopinath C, Rajvikram M (2018) A pathway to explore the hidden specialty in the design of fifteen level inverter in grid connected PV system. Comput Model Eng Sci 115(3):359–386

Coordinative Control of Tuned Fuzzy Logic and Modified Sliding Mode …

201

24. Rajvikram M, Saravanan M (2019) Efficient sliding mode pi controller for fault recovery in grid connected wind energy conversion system. J Electr Eng 19(1):44 25. Rajvikram M et al (2020) A state-of-the-art review on the drive of renewables in Gujarat, state of India: present situation, barriers and future initiatives. Energies 13(1):40 26. Rajvikram M (2019) Comprehensive review on India’s growth in renewable energy technologies in comparison with other prominent renewable energy based countries. J Solar Energy Eng 142:030801–030811 27. Rajvikram M et al (2018) A novel methodology of IoT implementation in energy management. Power Res 14(1):85–91

Sustainable Energy Technologies: Energy Resources for Portable Electronics—A Mini Review M. Malini , P. Sriramalakshmi , and M. Sujatha

Abstract Portable Electronic Devices (PEDs) are gaining more attention. Various portable electronics devices are used on a daily basis, and the best example for it is the smart phones due to its attractive features such as, light weight, compact size which enables it to be carried around easily. Owing to the continuous technological development, there is a drastic increase in the need for more energy sources. As the portable electronic devices evolve, the demand for new types of energy sources is growing. In the current situation, new developments are being made and the sustainable energy sources are now taking a front seat. A glimpse on various generation of computers, some of the developments, their benefits and issues are discussed in this research article along with few solutions. Keywords Generations of computer · Portable electronic devices (PEDs) · Energy resources · Sustainable energy sources

1 Introduction An exceptionally broad range of electronics is encompassed under the term electronics, which was originally used for the analysis of electron characteristics in the electron tubes. The research in the field of electronics has directed to the progress of essential devices like the lasers, IC’s, transistors, optical fibers and made it probable for the production of large types of industrial, electronic consumer goods and armed products. Indeed, it is said that the world is in the midst of an electronic revolution of the nineteenth century [1]. M. Malini · P. Sriramalakshmi (B) Vellore Institute of Technology, Chennai, Tamil Nadu, India e-mail: [email protected] M. Malini e-mail: [email protected] M. Sujatha Hindustan Institute of Technology and Science, Chennai, Tamil Nadu, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_17

203

204

M. Malini et al.

The history of electronics started when J. A. Fleming invented the vacuum diode in the year 1897 followed by the implementation of vacuum triode by Lee De Forest for the amplification of the electrical signals. The introductions of the tetrode and pentode tubes are invented followed to it. There are certain limitations such as bulky size and high power consumption for operation. Next, the invention of the junction transistor in the year 1948 earned the Nobel Prize and gained wide acceptance for usage in various electronic circuits when the germanium and silicon semiconductors replaced the vacuum tubes. There was a notable change in the electronic circuits when the invention of the Integrated Circuits (ICs) was witnessed. The integration of the electronic circuits into a single chip gives rise to low cost and compact size electronic devices. ICs with enlarged capabilities such as small-scale integration, medium-large scale and verylarge scale integration ICs were introduced during the period of 1958–1975.Further developments proceeded to the introduction of the Junction Field Effect Transistors (JFETs) and the Metal Oxide Semiconductor Field Effect Transistor (MOSFETs), developed during the years 1951–1958 and so it improved the device designing process making it more reliable. A robust development in digital ICs changed the overall architecture of the computers. These ICs are developed using logics such as Transistor-Transistor Logic (TTL), emitter coupled logic (ECL) and Integrated Injection Logic (I2L) technologies. These are employed in the fabrication technologies of P-type Metal Oxide Semiconductor (PMOS), N-type Metal Oxide Semiconductor (NMOS) and Complementary Metal Oxide Semiconductor (CMOS).The radical changes in components paved way for the microprocessors by Intel around 1969. Shortly, the analog integrated circuits were developed, leading to the introduction of operational amplifier [2].

2 Generations of Computer Evolution in electronics has changed various aspects drastically. One of the prominent evolutions is that of the computers. Starting from the simple calculating machines in 1940s to small sized, complex and powerful machines of today, the evolution of computers has been categorized in five generations, each having prominent technological developments in terms of their electronic components [3].

Sustainable Energy Technologies: Energy Resources …

205

Fig. 1 Illustrations of the five generations of computer [4–8]

2.1 First Generation The era of first generation of computers was from 1940 to 1956. The first generation computers were usually limited for scientific research [9]. Memory was an expensive aspect and cost a lot to run. Electronic time per calculation ranged from 0.1 to 1 ms. The main drawback of these computers is that it can solve one problem at a time [10, 11]. The first successful electronic computer called Electronic Numeric Integrated And Calculator (ENIAC) invented by J. P. Eckert and J. W. Mauchy, consist of nearly 20,000 vacuum tubes, 10,000 capacitors and 70,000 resistors, weighing over 30 tons and required a large room to house it [12]. Some other examples of first generation computers are: EDVAC, UNIVAC, EDSAC, Manchester Mark 1, Mark 1, Mark 2, IBM-709, 701, 700, 650 series. The UNIVAC is the first commercial computer owned by a businessmanin 1951—the US Census Bureau [13, 14].

2.2 Second Generation The replacement of the larger vacuum tubes by smaller transistors leads to the second generation of computing, from 1956 to 1963. William Shockley, John Bardeen and Walter Houser Brattain developed the first transistor at the bell laboratories in 1947 and initiated into widespread usage in computers around 1950s. Made of silicon, these transistors were less sensitive to temperature and did not burn up easily. The second generation computers were able to store the instructions into the memories [9].The initial types of these machines were meant for the atomic energy industry [10, 11]. The TX-0 was the first computer to use the transistors, followed by UNIVAC 1108, IBM 7070, IBM-7000, IBM-7094, CDC 3000, CDC 1604, CDC 3600, RCA

206

M. Malini et al.

501, Philco Transac S-1000, PDP1, PDP3, PDP5, PDP8, ATLAS and Honeywell 400 [12, 13].

2.3 Third Generation Integrated Circuits (IC’s), invented by Jack Kilby and Robert Noyce in 1958–1959, led into the introduction of the third generation computers from 1964 to 1971 [9]. Considered to be in use till the 1970s, even computers of today trace back deep roots to the third generation [10]. These were the first computers to interface using the operating systems and interfaced inputs given through keyboards and outputs displayed through the monitors [11]. A central program functioning to the computer memory, the operating systems enabled the development of multiprogramming. Mini computers were introduced with the miniaturization of the transistors put on silicon chips (semiconductors) during the 60s [12]. Some examples of the third generation of the computers are PDP-8, IBM 360, ICL 2900, IBM-307, PDP-11, IBM 370, CDC 7600 and PDP (Personal Data processer) II etc. [13].

2.4 Fourth Generation The fourth generation of computers came in around 1971 and lasted till 2010. Developed using the microprocessors, these computers fabricated using Large Scale Integration (LSI) and Very Large Scale Integration (VLSI) technology. The microprocessors more commonly known as the Central Processing Unit (CPU), resulted in fast Random Access Memory (RAM) [9]. The Intel 4004, the first microprocessor chip developed in 1971, and with the development of several operating systems such as MS-DOS and MS windows, the computers could be used for performing any logical and arithmetic functions and was a user friendly software. Problem-oriented fourth generation language (4GL) was used for the development of the programs. The first ever computer (IBM), designed in the year 1981 and the MacIntosh developed by the Apple, were specifically for household use and made computers an everyday product [10, 11]. Some of the other early computers to use microprocessors were the Altair 8800, IBM 5100, IBM 4341, IBM 3090/600, IBM AS/400/B60, Micral, DEC 10, CRAY 2, STAR 1000 and PUP 11 etc. The Graphics User Interface (GUI) finds advances in mouse, lap-top capabilities and the recent hand-held devices. Even though the fourth generation of computers is considered to have ended in 2010, today’s computers also use the microprocessors [12, 13].

Sustainable Energy Technologies: Energy Resources …

207

2.5 Fifth Generation The fifth generation computers use Super Large Scale Integrated (SLSI) and the Ultra Large Scale Integration (ULSI) chip technology starting from 2010 to till date in development. Mainly focusing on developing computers that can respond to natural language and the surroundings by making use of different kinds of sensors, these computers work on the basis of Artificial Intelligence (AI), which has been made possible by using superconductors and parallel processing which is much faster than serial processing [9, 10]. Some of the well-known examples of AI in computers are the IBM’s Watson, Apple’s Siri (for iPhone) and Microsoft’s Cortana (for Windows 8 and 10). The laptops, desktops, notebooks, ultra-books, chrome-books etc. are also examples for the same [12]. These computers have quality in intelligence; make default assumptions and have decision making capabilities. Ultimately this generation of computers are expected to be intelligent as humans in the future and ‘robots’ are a good example for this [10]. The developments in the field of AI have bought forth efficient technologies such as voice recognition and facial recognition. Despite these developments quantum computation, molecular, and Nano technology, provide much more scope for development in the future, where the research and development continue [11, 14] (Table 1). Table 1 Comparative table on generations of computer [10–13, 15, 16] 1st generation

2nd generation

3rd generation

4th generation

5th generation

Electronic component

Vacuum tubes

Semiconductor chips and transistors

Integrated circuits (ICs)

Microprocessors (LSI, VLSI)

Artificial intelligence (AI), (ULSI)

Input

Punch cards and paper tapes

Punch cards

Keyboard and mouse

Keyboard and mouse

Speech input, tactile input

Output

Printed reports

Printed reports

Printed reports and Monitors, printers monitor and speakers

Graphics displays, voice responses

Memory

Magnetic drums and tapes

Magnetic core, tapes and disks

Magnetic core, tapes and disks

RAM, ROM, cache memory (extendable to magnetic disk, floppy disk & optical disk-CD, DVD)

Optical disk and magnetic disk

Programming language

Machine and assembly language

Assembly language and high level language (FORTAN, COBAL, BASIC)

High level (PASCAL, COBAL, BASIC, C, etc.)

All high level Artificial languages (C+ + , intelligence (AI) KL1, RPG, SQL) languages (LISP, PROLOG)

Operating time

Milliseconds

Microseconds

Nanoseconds

Picoseconds

Fractions of picoseconds (continued)

208

M. Malini et al.

Table 1 (continued) Size

1st generation

2nd generation

3rd generation

4th generation

5th generation

Very big (size of room)

Reduced

Small (Mini computers disk sized)

Smaller (Micro)

Very small (Micro)

Lesser

Low

Cost

Very high

High

Lesser

Weight and portability

Heavy and non portable

Heavy and not easily portable

Heavy but portable Light and portable Very light and very easily portable

Efficiency (instructions per second)

Hundred

Thousand

Million (time sharing and multi programming)

Heat generated

Very high (AC High (AC required) required)

Less (AC required) Negligible (AC required in case of ICs)

Negligible

Energy consumption

Very high

Less

Low

High

Tens of millions

Less

Billions (parallel processing)

Reliability

Very less

Less

High

High

Very high

Maintenance (failure rate)

Seconds

Days

Weeks

Months

Year

3 Portable Electronics A Portable Electronic Device (PED) are typically consumer electronic devices capable for communications, data processing, and data storage and/or utility [17]. This field includes a range of devices and gadgets such as mobile phones, notebook computers, tablets, laptop, computers, etc. [18–20].

Fig. 2 Major electronic components used in generations 1–5 of electronics [21–25]

Sustainable Energy Technologies: Energy Resources …

209

3.1 Conventional Energy Resources for Portable Electronics and their Issues Recent trends in the portable electronic devices are favoring processors with highperformance, larger displays and storage, enhancement in the quality of the audio and the video, increased speed in wireless networking and overall a slim and lighter weighing package. It leads to more demand on the portable energy sources. The rechargeable batteries used in these portable electronic devices have lower capacity; size and longer life are the main parameters for the design of the system. With the increasing need for the smaller and improved sources of power for these consumer portable electronic devices, popular batteries such as the Li-ion batteries are already reaching their limit in the ability to provide the necessary amount of energy (operational time). The batteries rely on the electrochemical reactions happens within the cell, which does not follow scaling laws similar to the solid-state electronics, which benefit continually from the device level scaling and continued miniaturization of the package. Also limited materials are suitable for the rechargeable energy storage. The specific capacity depends on the crystallographic and morphological nature of the material being used. Most of the commercial battery technologies limit their specific capacities. The battery technology has become a critical bottleneck in the determination of the overall performance of portable electronic devices. The efficient management of the power, the battery architectures and the integration of low power drain components are turned important aspects in the system design of electronics. The increasingly broadening gap requires a paradigm shift in the way towards PES [26–28].

3.2 Available Options Owing to the ever increasing demand for the power sources for portable electronics, significant attention is given by the researchers in the field of development of the ambient energy sources. The last decade has witnessed the shrinking in volume of the electronic devices. Internet connectivity is enabled to carry the knowledge of the world in our pockets. Computation has become intuitive enough for kids due to the touch based tablets. Despite the huge developments in other fields, the batteries seems to be lagging behind. While the clock speeds in the device are measured in GHz, the life of the batteries is still measured in hours. Developments in this field are promising to take the batteries’ life to last for weeks. Running the devices in standby mode and operation times for longer periods require the batteries to be charged by some external energy sources regularly. Generally conventional methods of charging, discharging mechanism, rectification and filtering cause the mobile batteries different types of losses resulting in inefficient utilization of power [29]. Traditionally intended for applications of larger scale, the fuel cell could fit in potentially for the recent requirements due to their high-energy content of the fuels.

210

M. Malini et al.

Thus different types of fuel cell technology are being considered to provide replacement to the batteries for portable devices. A recent and novel technology in nanoscale fuel cell could give advanced powering solutions for portable electronic devices [28]. A new approach to improve the performances of electrochemical energy sources is to replace carbon atoms with nitrogen atoms in the crystal lattice of the super capacitor and developed novel method of capacity enhancement based on the modification of the carbon lattice with the help of plasma. This research could assist in the next generation of power sources for portable electronic devices. Next, the super capacitor, a chemical current source distinguished by its high charging and discharging rates compared to a battery. Porous materials such as porous metals or carbon are used for super capacitors. Generally metals make the source heavier. A group of scientists led by Skoltech Senior Researcher Dr. Stanislav Evlashin proposed a simple way of increasing the electrochemical performance of the super capacitors. It gives better understanding into the amalgamation process of nitrogen. The researchers performed the experiments using vertically oriented graphene sheets to make carbon nanowalls, replacing some of the carbon atoms with nitrogen atoms using carbon structure treatment by plasma. The outcomes of the study are an important footstep for formation of new energy sources [30].

4 Energy Resources for Portable Electronics The batteries play the key role in electronic devices which convert chemical energy directly to electricity. The lithium ion and the lead acid batteries are the most commonly used. The production of lead acid batteries is very easy and there is mass production of these batteries around the world, making them the most cost effective. The lead acid battery is used largely in higher power applications and its low cost provides an added advantage. The battery is charged by the reverse process which creates back the original lead and acid. On an average, the lifespan of these batteries is 1000–2000 cycles at 70% discharge. Renewable energy systems of today require batteries for its operation and for implementation of its different charging techniques. Two different essential operations of battery sources are the storage of the energy produced and smoothen the energy being produced. These lead acid batteries widely used in wind and photo voltaic applications. Solar panels produce electricity during sunlight hours but need batteries to store the electricity for use later. Thus the effectiveness of these systems is getting limited to the efficiency of the batteries used to store and release energy. Likewise, the electricity generated by the wind turbine is affected by the wind speed which changes its speed almost every instant. Batteries are required for the stabilization of the inconsistent energy surges. Consumable power can be achieved from the system only when the batteries can stabilize the energy of the system effectively. New technologies are developed to find replacement for the lead acid technology and to make the systems cost effective [31–33].

Sustainable Energy Technologies: Energy Resources …

211

4.1 Sustainable Energy Resources With the ever increasing demands for energy sources and its longer durability, sustainable energy resources seem to be the need of the hour. The PV and wind are the potential energy sources. Being environmental friendly, renewable and long lasting make them potential candidates for being used in portable electronic devices. Portable RES are the standalone devices that are powered by energy alternatives such as the wind and the solar energy. Being lightweight, and safer due to the absence of chemicals generally used in batteries (such as lead), and environment friendly make them a better alternative than the widely used chemical and short lived methods. Though traditional appliances require regular refueling and contribute to local air pollution due to the emission of sizeable toxins. Nevertheless, the solar power remains the dominating source in the portable renewable energy market, followed by the wind power. The availability of thin film solar cells, especially second generation solar cells include the cadmium telluride, amorphous silicon and copper indium gallium selenide solar cells result in the rise of portable solar power [34].

4.2 Features of Solar Energy Solar energy is an upcoming popularly used source of renewable energy. Making use of solar panels or photovoltaic systems, large silicon based tiles is placed in areas of unobstructed sunlight. These tiles take in the solar energy and convert it into electricity. These are of two types basically, “on-grid” and the “off-grid”. The grid refers to an interconnected network of houses or buildings to a power plant. The grid type photovoltaic systems specifically use batteries. The on-grid systems are connected to the network and the required power is made available from elsewhere. The off grid systems are totally disconnected from a power plant and all the energy used in the system is produced on its own. In this case, the solar panels produces the off grid energy. Usable energy is produced only during daylight, and stored in batteries for use in later times. The off grid photovoltaic system, have the solar tiles as the primary source of energy while the battery acts as a backup source. The solar energy is used when it is available and when it is not the battery power is used. The limits of the batteries do not allow all the energy harnessed by the tiles to be used. The size and the amount of electricity used dependents on the electricity used within the system. The short lifespan of the batteries make them the most expensive components in the system configuration. Certain charging techniques are used in order to increase the lifespan of the batteries. But, these techniques limit the efficiency of the system. Though lead acid batteries are used commonly due to its mass production, they are not the ideal one. Thus maximizing the battery life, limits the efficiency of the other parts and the replacement with every 1000–2000 cycles makes them quite expensive. Massive improvements in battery technology in

212

M. Malini et al.

Table 2 Battery classifications for RES [32] Battery category Advantages

Disadvantages

Common uses

Lead acid

• Low cost • Mass produced

• Heavy-harmful chemicals • High maintenance • Short life

• PV panels • Wind applications • Standby systems

Lithium ion

• Compact size and weight • Expensive • Increased energy • Difficult to produce capacity • Long life

• Small electronics

Ultra capacitor

• Long life • High power • Quick charging

• None yet

• Expensive • Difficult to produce

the recent years have focused on more power in smaller space, this is specifically advantageous for low small handheld electronic devices, but the larger applications still face problems. The lead acid batteries still use liquids inside them to create electricity, and these chemicals are dangerous and could damage the environment if not disposed of correctly. Scaling up newer battery technologies to the power of lead acid batteries are not worth the price. The lithium ions being specifically designed for the larger applications, other battery technologies are being tested in renewable energy application. The ultra-capacitors have much longer life-span than the lead acid batteries, and do not rely on chemical reactions. They can be discharged completely and can be stored in any state without any problems. These can be charged much faster as the charges flow much quickly in and out. Though potential for a more efficient and long lasting batteries, and have the ideal characteristics for renewable application, mass production and cost effectiveness will be required before they are widely implemented. Table 2 lists the advantages, disadvantages and common uses for major battery types [32].

4.3 Upcoming Options in Sustainable Energy The need of longstanding and efficient power sources have led the researchers to turn towards the sustainable energy resources. Some upcoming options in the field are discussed below. Hybrid renewable energy power systems pose a possible long term power solution for transportation. The systems consist of PV, wind, fuel cells, electrolyzers, batteries, super capacitors and other power devices. These systems set up with power electronics to handle low, high and variable power. Here the power electronics play a key role in modulating and managing the power as per the requirements [35, 36].

Sustainable Energy Technologies: Energy Resources …

213

A system proposed in [37], combines the PV and the wind energy together with a battery storage system that can provide supply to the low power device. In this system, if any surplus energy is produced from the renewable sources, then this excess energy is stored in the battery. If the energy produced is insufficient, then the energy is provided by the battery. The modulating strategies used in converter ensure the optimized use of the sources [38]. The usage of energy harvesting is mostly cited due to battery maintenance and replacement. A primary power source is required to charge them as the rechargeable batteries which are secondary power source. Energy harvesting sources such as solar cells could be used in applications where rechargeable batteries are used. Where higher outputs are required from hours to days, fuel cells and ultra- capacitors are potential options. Portable, low power communications and computer devices are one of the biggest markets for energy harvesting. Portable medical devices are other good alternatives for energy harvesting. The exploitation of body heat as thermal energy source makes thermoelectric generators. It proves to be beneficial for medical devices and biomedical implants. Energy can be harvested from body heat and is not limited to devices applied to the skin. Active implantable devices are located under the skin for applications such as cardiac pacemakers, muscle stimulators, neurological stimulators or Cochlear implants. Researchers are involved in finding solutions to design mini solar panels into portable handsets to extend the standby time. And considering the projections in the worldwide market for portable electronics (like mobiles), the sustainable energy could become “the next big thing” in power [38, 39].

5 Conclusion The development of various Portable Electronic Devices (PEDs) and their usability of in today’s scenario is well discussed. With the flourishment of more and more devices, the importance of energy resources needed for their operation is reviewed. The growth of the new types of sustainable energy sources and the existing issues with conventional energy resources such as battery capacity, charging time etc., are debated and alternative options like nano fuel cells and super capacitors are introduced. Energy harvesting techniques as a supplement to battery is suggested and parallelly operated miniature solar cells along with battery sources is discussed for further improvement.

References 1. Scace RI Electronics. Encyclopædia Britannica (February 15, 2016/02/15) 2. Brief history of electronics and its development, https://www.elprocus.com/know-about-briefhistory-of-electronics-and-their-generations/

214

M. Malini et al.

3. Andreea I (2015) Categories and generations of computers. Euro J Comput Sci Inf Technol 3(1):15–42 (European Centre for Research Training and Development) 4. Generations of computer, https://semesters.in/generation-of-computers. Accessed on 06 June 2015 5. Computer generations, https://ccsindian.in/2017/03/27/computer-generation. Accessed on 07 Mar 2017 6. Basic computer, https://generalnote.com/Basic-computer/Third-Generation-of-computer.php 7. Bdstall, https://www.bdstall.com/details/4th-gen-pentium-dual-core-500gb-16-desktop-com puter-15958/. Accessed on 2016 8. Computer history, https://computerhistoryforyou.weebly.com/4th--5th-generations.html 9. Generations of computer, https://www.physics-and-radio-electronics.com/computer-basics/ generations-of-computer/generations-of-computer.html 10. Generations of computer and their characteristics, https://www.vidyagyaan.com. Accessed on 12 July 2018 11. Five generations of computer, https://btob.co.nz/business-news/five-generations-computers/. Accessed on 15 Nov 2015 12. Computer Hope, Generations of computer, https://www.computerhope.com/issues/ch001921. htm. Accessed 30 June 2019 13. Generations of computer, https://www.geeksforgeeks.org/generations-of-computer/. Accessed on 09 May 2019 14. Thefreedictionary.encyclopedia2. https://encyclopedia2.thefreedictionary.com/computer+gen erations. Accessed 30 Nov 2019 15. Sharma P (2015) Generation of computers: the evolution. Yourarticlelibrary 16. Beal V (2018) The five generations of computers. Webopedia 17. PRNewswire (2018) Global portable electronics market research report—forecast to 2023 18. Duquette A Fact sheet—portable electronic devices aviation rulemaking committee report (2013/10/31) 19. Robert Kebel I, Thiemo Stadtler I (2017) Progress in usage of portable electronic devices on aircraft, an overview. In: Proceedings of the 2017 international symposium on electromagnetic compatibility—EMC Europe, IEEE, Angers, France. https://doi.org/10.1109/emceurope.2017. 8094746 20. Turiak M, Novák-Sedláˇcková A, Novák A (2014) Portable electronic devices on board of airplanes and their safety impact. In: Mikulski J (ed) Telematics—support for transport. TST 2014. Communications in computer and information science, vol 471. Springer, Berlin, Heidelberg 21. Vacuum tubes, https://www.circuitstoday.com/vacuum-diodes. Accessed on 17 Aug 2009 22. Electrical Fundamentals Vacuum Triodes, https://www.angelfire.com/electronic/funwithtubes/ Basics_04_Triodes.html. Accessed on 04 Mar 2004 23. Characteristics and Working of P-N Junction Diode, https://teknogenius.blogspot.com/2014/ 10/characteristics-and-working-of-p-n.html. Accessed 24 Oct 2014 24. Timer, https://www.circuitstoday.com/555-timer. Accessed 28 Aug 2018 25. Hrishikesan S Functional block diagram of 8085 microprocessor, electronics and communication, 30 June 2019 26. Chalamala BR (2007) Portable electronics and the widening energy gap. Proceedings of the IEEE, IEEE 95(11):2106–2107 27. Home Electronics, https://smarterhouse.org/appliances-energy/home-electronics 28. Dyer CK (2004) Fuel cells and portable electronics. In: 2004 symposium on VLSI circuits, digest of technical papers. Honolulu, HI 29. Khosla R, Choudhary R, Tiwari NK, Chandra D, Yadawa AK (2014) Review of energy sources for futuristic mobile electronic devices. Int J Adv Res Electric Electron Instrum Eng 3(7):10638–10646 30. Evlashin SA et al Scientists find a way to increase the capacity of energy sources for portable electronics (phys.org, 28 May 2019)

Sustainable Energy Technologies: Energy Resources …

215

31. Zhou C, Yang Y, Sun N (2018) Flexible self-charging power units for portable electronics based on folded carbon paper. Nano Res 11(8):4313–4322 32. Slinger K (2015) Electrical batteries for renewable energy. In: Electrical and computer engineering design handbook 33. Jung Y, Chang T, Zhang H et al (2015) High-performance green flexible electronics based on biodegradable cellulose nanofibril paper. Nat Commun 6:7170 34. Tan M (2015) Portable renewable energy systems present a convenient and green alternative to generate electricity on the go. Frost & Sullivan 35. Mardani A, Jusoh A, Zavadskas EK, Cavallaro F, Khalifah Z (2015) Sustainable and renewable energy: an overview of the application of multiple criteria decision making techniques and approaches. In: Torretta V (ed) Sustainability, vol 7, pp 13947–13984. ISSN, 2071-1050. https:// doi.org/10.3390/su71013947.MDPI 36. Spiegel C (2018) Power electronics for renewable energy systems. Fuel cell store 37. Brush L (2007) Portable devices: the benefits from energy harvesting. EDN Network 38. Krishnan JR, Suryamol MS, Varghese A, Ramesh P (2017) Design of a sustainable power source for portable electronic applications. Int J Adv Res Trends Eng Technol (IJARTET) 4(Special Issue 18):107–113 39. Ellsmoor J (2018) 6 Renewable energy trends to watch. In: 2019 Forbes

Power Electronics

Simulation and Analysis of a Voltage Control Strategy for Single-Stage AC-AC Converter A. Jamna, R. Sujatha, and V. Jamuna

Abstract Matrix converters (MC) are essentially four-quadrant forced commutated converter and are formed by connecting bidirectional switch cells in a matrix form. Factors like single-stage transformation of voltage and frequency, bidirectional power flow capability, sinusoidal input and output voltage, more compact size, low weight, and DC link-free architecture make this converter more attractive than a standard twostage converter. MCs can be used in applications such as drive systems, frequency converters, wind energy conversion system, and power supply units. In this chapter, the simulation of the MC system and its control strategy under different operative conditions has been discussed. The simulation helps to gain insights into the effect of different parameters on the converter performance. Keywords Single-stage AC to AC converter · Matrix converter · Voltage control · MATLAB

1 Introduction MCs are generally classified into two types, single-phase matrix converters (SPMC) consisting of four four-quadrant switches and a three-phase MC consisting of nine four-quadrant switches. Any given switch in MC is capable of conducting current in both directions, while at the same time being capable of impeding forward and reverse voltages. In 1981, Alesina et al. kick started work on MCs. They had familiarized the ACAC converter as a matrix of four-quadrant switches, and hence named it as matrix

A. Jamna (B) · R. Sujatha (B) St. Joseph’s College of Engineering, Chennai 600119, India e-mail: [email protected] R. Sujatha e-mail: [email protected] V. Jamuna Jerusalem College of Engineering, Chennai 600100, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_18

219

220

A. Jamna et al.

converter [1, 2]. Baharom et al. explained the feasibility of MC to operate as a rectifier in buck and boost mode [3, 4]. Deivasundari et al. analyzed and developed the application of SPMC as rectifier and inverter [5]. Idris et al. recommended a lucid commutation strategy for the SPMC. The proposed strategy included the required freewheeling operation similar to those available in other converter topologies [6]. Minh-Khai Nguyen et al. proposed a single-phase Z-source buck-boost MC in which both the frequency and the voltage could be stepped up and down. The duty ratio of the power electronic switches provided in the Z-network was controlled to obtain the required output voltage [7]. The design and implementation of a DC-AC converter using SPMC topology were implemented with bidirectional capabilities. To synthesize the output voltage, PWM technique was employed [8]. Rahimi Baharom et al. proposed a topology of the single-phase MC to function as a rectifier. A commutation strategy of the converter for R and RL loads was discussed [9]. Taha Ahmad Hussein proposed an improved SPWM control strategy for the reduction of harmonics from the output of cycloconverter based on SPMC [10]. MC transforms the input voltage, V i (t) through these four-quadrant switches to the desired output voltage V o (t), in accordance with the pre-calculated switching angles. The output voltage is a function of the input voltage and the modulation matrix. The desired voltage at any particular frequency is generated by appropriately selecting the switching pattern. The general equation for the output voltage of an MC can be written in terms of input voltage V i (t), modulation matrix m(t), and output voltage V o (t) as, V0 (t) = m(t)Vi (t)

(1)

Figure 1 shows the circuit diagram of a single-phase MC. It consists of a 2 × Fig. 1 Circuit diagram of a single-phase MC

Simulation and Analysis of a Voltage Control Strategy …

221

2 matrix of bidirectional switch cells S1, S2, S3, and S4 that connect the input and output lines. Each individual switch cell is capable of conducting current in both directions and also capable of blocking both polarities of voltages. The output voltage with any frequency can be obtained by applying switching pulses of appropriate sequence. Two important rules have to be followed to provide the safe operation of the MC. They are, • No two input lines should be shorted so as to avoid any short circuit at the input. • No output line should be open circuited so as to avoid any open circuit at the output. With the rules of safe commutation and with respect to Fig. 1, the operation of an SPMC can be explained in four modes. Mode 1, with V i (t) positive and V o (t) positive, (S1a and S4a ) are in ON state. Mode 2, with V i (t) positive and V o (t) negative, (S2b and S3b ) are in ON state. Mode 3, with V i (t) negative and V o (t) positive, (S2a and S3a ) are in ON state. Mode 4, with V i (t) positive and V o (t) negative, (S1b and S4b ) are in ON state. Therefore, the output of the converter with any frequency can be achieved by appropriately selecting these four modes. Depending upon the output frequency, these modes are repeated a number of times. Figure 2 shows the modes of operation of SPMC.

Fig. 2 Modes of operation of SPMC

222

A. Jamna et al.

1.1 Simulink Model of SPMC Model of an SPMC as shown in Fig.3 is simulated using MATLAB/Simulink with resistive and inductive loads. The desired output of the SPMC is obtained by appropriately selecting the ON/OFF states of the bidirectional switches. SPWM technique is used to generate the pulses for the switches, and the performance of the system is compared for various frequency values. The circuit is designed for an input voltage of 230 V, 50 Hz and an output of 110 V, at various frequencies. For the control of the converter, SPWM method is adopted. The performance of the converter is analyzed and compared for both the control methods (Table 1). For a purely resistive load, the output voltage does not has any voltage spikes but large spikes during switch transitions are produced at the output of the converter when inductive loads are used. These spikes increase the switching stress and may damage the switches in use. Therefore, in this work, a new and efficient commutation switching method is introduced to eliminate these spikes. In the proposed method, the sequence of switching is arranged in such a manner that the bidirectional switches itself act as the freewheeling path. Table 2 shows the modulation of switches.

Fig. 3 MATLAB/Simulink model of SPMC

109.3

108.7

106.9

106.1

100

200

400

2.86

2.53

2.04

1.52

4.83

4.96

5.035

5.467

2.86

2.53

2.04

1.52

50.91

53.53

62.01

71.3

Magnitude of fundamental voltage (V)

Harmonic content in the output current (%)

Inductive load Magnitude of fundamental current (A)

Magnitude of fundamental voltage (v)

Harmonic content in the output voltage (%)

Resistive load

50

Output frequency (Hz)

Table 1 Performance analysis with resistive and inductive loads

78.82

74.28

36.94

34.3

Harmonic content in the output voltage (%)

1.9

4.3

6.52

9.5

Magnitude of fundamental current (A)

38.36

34.77

30.08

28.7

Harmonic content in the output current (%)

Simulation and Analysis of a Voltage Control Strategy … 223

224

A. Jamna et al.

Table 2 Modulation of switches for an output frequency of 400 Hz Input frequency

Output frequency

Mode

Switches ON

Commutation switch ON

50 Hz

400 Hz

1

S1a and S4a

S2a

2

S2b and S3b

S1b

1

S1a and S4a

S2a

2

S2b and S3b

S1b

1

S 1a and S4a

S2a

2

S2b and S3b

S1b

1

S 1a and S4a

S2a

2

S2b and S3b

S1b

3

S2a and S3a

S1a

4

S1b and S4b

S2b

3

S2a and S3a

S1a

4

S1b and S4b

S2b

3

S2a and S3a

S1a

4

S1b and S4b

S2b

3

S2a and S3a

S1a

4

S1b and S4b

S2b

For inductive load, large voltage spikes are produced at the output during the switching transitions. Hence, a safe commutation strategy is developed to reduce these spikes. The magnitude of fundamental value of voltage and current is improved with safe commutation strategies, whereas THD values are reduced with safe commutation switching techniques. Also, the dv/dt stress on the switches is reduced considerably. Performance comparison of SPMC with and without safe commutation strategy is given in Table 3.

2 Voltage Control of SPMC The main feature of an MC is its capability to transform the magnitude as well as the frequency of the AC input voltage into an AC output voltage of required magnitude and frequency as an “all-silicon” solution. The magnitude of the output voltage depends highly on the types of load, input supply disturbances, sudden connection, or disconnection of load. The converter system should be capable of tackling the variation in the output voltage due to these reasons and to maintain a constant magnitude and frequency AC voltage at its output terminals. Therefore, simulation of SPMC is executed and behavior of the system is studied for various load conditions. A source disturbance is created to check the capability of the system to cope with change in input. By using the open-loop simulation parameters, a voltage control loop is designed and simulated. The expediency of the

71.3

62.01

53.53

50.91

100

200

400

78.82

74.28

36.94

34.3

1.9

4.3

6.52

9.5

38.36

34.77

30.08

28.7

440

334

330

30

100.9

101.4

102.7

103.0

22.99

17.4

13.3

8.7

Magnitude of Harmonic fundamental content in Voltage (v) the output voltage (%)

Stress on the switches (V/ns)

With safe commutation Harmonic content in the output current (%)

Magnitude of Harmonic fundamental content in Voltage (v) the output voltage (%)

Magnitude of fundamental Current (A)

Without safe commutation

50

Output frequency (Hz)

Table 3 Performance comparison of SPMC with and without safe commutation strategy

2.5

3.2

7.51

12.5

Magnitude of fundamental Current (A)

17.9

11.4

10.2

7.9

Harmonic content in the output current (%)

0.158

0.133

0.127

0.039

Stress on the switches (V/ns)

Simulation and Analysis of a Voltage Control Strategy … 225

226

A. Jamna et al.

Fig. 4 Open-loop Simulink model of SPMC with disturbance

SPMC control is tested by observing the converter’s response to changes in input parameters.

2.1 Matrix Converter with Source Disturbance The model of the MC with a source disturbance is shown in Fig. 4. The converter is simulated with resistive–inductive load condition. An input of 230 V, 50 Hz is applied to the converter. For creating an output voltage at 400 Hz, with an input voltage at 50 Hz, during the positive half cycle of the input voltage, the switch pairs (S1a, S4a) and (S2b, S3b) are operated and during the negative half cycle of the input voltage (S2a, S3a) and (S1b, S4b) pairs are operated. The switching pulses generated for an output voltage of 110 V, 400 Hz driving inductive load are illustrated in Fig. 5. The output of the converter may be affected by number of parameters such as disturbance in input supply and instant disconnection of load. During these abnormal conditions, the MC should be able to maintain its output voltage. To check the feasibility of the system to cope with any change in input supply, two circuits breakers are introduced at the input in the simulation circuit. Breakers are connected in such a way that the input voltage is increased for a short duration of time from the constant value. A disturbance at the input is created at 0.5 s. Figure 6 shows the disturbance

Simulation and Analysis of a Voltage Control Strategy …

227

Fig. 5 Switching pulse generated for obtaining 110 V/400 Hz with safe commutation

AC input supply in volts

400

200

0

-200

-400

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

time is secs

Fig. 6 Disturbances applied to the input of the system

created. At a time interval of 0.5 second, the breaker one is opened and breaker two is closed. The input voltage is increased from 230 to 280 V. The resultant output voltage wave form is shown in Fig. 7. An input disturbance is created at 0.5 s and it is found that the output voltage of the converter is increased from its constant output value at the 0.5 s. Hence, a controller has to be introduced to keep the output voltage constant. In this chapter, the performance of the converter is analyzed using P controller and a PI controller and is compared. Four important parameters to be considered while designing closed-loop system are the rise time, overshoot, settling time, and steadystate error. These four parameters can be improved by properly tuning the control parameters, proportional gain (K p ), integral gain (K i ), and differential gain (K d ).

228

A. Jamna et al.

Fig. 7 Output voltage waveform of an open-loop system with disturbance

Ziegler and Nicholas method has been adopted for the design of the controller. The steps to be followed for a controller are: • To determine the open-loop characteristics that has to be improved. • Calculate the K p , K i , and K d depending on the system requirement. Increase in K p decreases the rise time, K d reduces the overshoot and settling time and K i eliminates the steady-state error. Based on the transient characteristics, Zieglar proposed a rule to determine the values of K p , K i , and K d . He made use of an S-type unit reaction curve. This curve can be defined by the parameters, namely delay time and time constant. These parameters are obtained by drawing a tangent line at the inflection point of the reaction curve. The inflection point of the drawn tangent line with the time axis and steady-state value, L and T, can be obtained. The closed-loop system is designed using P controller, and the control parameters are calculated by using Ziegler and Nicholas method. The equations used for the design of the controllers are:   1 c(s) = K P 1 + Ti S c(s) = K P + Ki =

Ki S

KP Ti

(2) (3) (4)

The transfer function of the process G(s) can be written as, G(s) =

K e−s L Ts + 1

(5)

An MC system with a simple PI controller that controls the output voltage even with any input disturbance has been developed as shown in Fig. 8.

Simulation and Analysis of a Voltage Control Strategy …

229

Fig. 8 Simulink model of MC system with PI controller

An input voltage of 230 V, 50 Hz supply is applied to the MC. The MC drives a resistive–inductive load which is connected to the MC after the output filter stage. The maximum voltage and frequency required for the system are 110 V rms and 400 Hz, respectively. An input disturbance as shown in Fig. 6 is created to test the performance of the system. The rms output voltage waveform of MC system with PI controller is shown in Fig. 9, respectively. The value of proportional and integral gain is calculated using Zieglar–Nicholas method by finding the values of L and T from the S-curve. From the output voltage wave form, the rise time, settling time, delay time, and percentage overshoot are calculated (Table 4).. 120

output rms voltage

100 80 60 40 20 0

0

0.1

0.2

0.3

0.4

0.5

time in secs

0.6

Fig. 9 Output voltage waveform of the system with PI controller

0.7

0.8

0.9

1

230 Table 4 Comparison of MC with P, PI controllers

A. Jamna et al. Performance parameter

P controller

PI controller

Rise time (s)

0.007

0.006

Delay time (s)

0.002

0.001

Settling time (s)

0.3

0.03

Over shoot (%)

8.8

4.4

Steady state error (V)

3.3

1.3

3 Conclusion The MCs are the right solution in applications which demand the low weight, less space, high reliability and the performance of the converter is a major concern. The input displacement angle control as well as the bidirectional power flow capability of the MC makes it suitable for applications where regeneration is demanded. In this chapter, the expediency of the SPMC control is tested by observing the converter’s response to changes in input parameters. Moreover, the simulation helps to gain insights into the effect of different parameters on the converter performance. A closed-loop system with P and PI controller is simulated using MATLAB, and the performance of the system is analyzed.

References 1. Alesina A, Venturini M (1981) Solid state power conversion: a fourier analysis approach to generalized transformer synthesis. IEEE Trans Circuits Syst CAS-28(4):319–330 2. Alesina A, Venturini M (1989) Analysis and design of optimum amplitude nine-switch direct AC-AC converters. IEEE Trans Power Electron 4:101–112 3. Baharom R, Hamzah R, Hamzah KS, Saparon MK (2009) SPMC operating as buck and boost rectifier. IEEE AE Con, pp 3338–3342 4. Baharom R, Hasim Hamzah MK, Omar (2006) A new single-phase controlled rectifier using SPMC. IEEE PECon, pp 453–458 5. Deivasundari P, Jamuna V (2011) Single phase matrix converter as an all silicon solution. In: 1st international conference on electrical energy systems, IEEE 6. Idris Z, Hamzah MK, Saidon MF (2006) Implementation of SPMC as a direct AC-AC converter with commutation strategies, IEEE PESC, pp 2240–2246 7. Nguyen M-K, Jung Y-G, Lim Y-C, Kim Y-M (2010) A Single phase Z-source buck-boost matrix converter. IEEE Trans Power Electron 25(2):453–463 8. Noor M, Zaliha S, Hamzah MK, Baharom R (2007) A new single-phase inverter with bidirectional capabilities using SPMC, IEEE PESC, pp 464–470 9. Baharom R, Hamzah MK (2007) A new single phase controlled rectifier using single phase matrix converter topology incorporating active power filter. In: Electric machines & drives conference-IEMDC, IEEE 10. Hussein TA (2009) Voltage spikes and harmonics reduction for inductive load single phase matrix converter using MATLAB. In: ICCEE, pp 219–224

Implementing PV Energized SRSPM-Based Single-Phase Inverter for Induction Motor P. Abirami, M. Pushpavalli, and P. Sivagami

Abstract This research work explains in detail about the SRSPM technique along with solar energy. The drawback of the existing system is that it leads to high cost because of fuel cells and implements two isolated DC–DC converters in it which leads to high switching losses. Thus, to overcome the above said drawbacks, a singlereference six-pulse modulation technique along with solar power is implemented to improve the efficient usage of electrical vehicles. So, in this research work, solar energy is used to run the motor which is the renewable source energy. The PV panel generates DC supply, and a pulsating DC is produced by the hybrid technique SRSPM which is used to improve the efficiency of the vehicle. Here, the HF modulation is adopted to obtain compact, low cost, and lightweight system. Thus, the low voltage signal is amplified to the required three-phase voltage signals by means of a multistage inverter. HF inverter consists of a front-end DC–DC converter and three-phase PWM inverter. Here, two different modulations are implemented. SRSPM is used to control front-end full-bridge DC–DC converter to produce HF pulsating DC voltage having six-pulse information, and the average switching frequency and switching losses are reduced by balanced three-phase voltage produced by implementing 33% modulation technique of three-phase inverter. Thus, the proposed system simulation is carried out for three-phase supply and experimental setup is implemented for a single-phase induction motor. Keywords Fuel vehicles · Electrical vehicle · SRSPM · Induction motor first section

1 Introduction The fuel vehicles are used for the transportation purpose. Nowadays, the electrical vehicles are implemented to get a pollution-free environment and also to reduce the cost of the vehicle. In that electrical vehicle, batteries are used to run the motor. To run P. Abirami (B) · M. Pushpavalli · P. Sivagami AP/EEE Sathyabama Institute of Science and Technology, Chennai, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_19

231

232

P. Abirami et al.

a vehicle motor, enormous amount of energy is required. The high amount of energy requires higher capacity battery. So, the size of the battery required for the electrical vehicle is high. Because of the above said fact, the vehicle size and weight of the vehicle are higher than that of the normal fuel vehicle. Also, it is required to perform the periodic maintenance of the batteries or to charge the batteries on the daily basis for routine usage. In our modern life, vehicles play an important role. Most of the vehicles are running based on the fuel like diesel, petrol, and LPG gas [1–5]. Since these fuels are in great demand in our day-to-day life, we are in a position to utilize an alternate source in place of these fuels. Recent trends in electrical engineering provides a solution for this critical situation by improving the renewable energy resources [6–8]. Most widely used renewable energy resources are solar energy and wind energy. For above, the electrical vehicles are introduced and used widely. In the electric vehicles, the mechanical rotation is achieved by converting the electrical energy to mechanical energy. Internal combustion engine (ICE) is used in conventional vehicles. Because of its poor performance, ICE converts only 10–15% of the fuel energy to traction. This performance can be improved to 30–40% by implementing hybrid electric vehicles (HEV) [8–12]. By implementing HEVs, emission of CO2 can be eliminated to a certain level. But, by using electric vehicles (EVs), CO2 emission can be eliminated completely. Energy storage is a primary drawback of EVs, and also the availability of storing options is minimum. Another drawback is, even if storing devices are available, the energy storage device has to be charged quickly. Fuel cell vehicles (FCVs) are the solution for this issue but it is dependent on foreign oil.

1.1 SRSPM Hybrid Technique The SRSPM stands as single-reference six-pulse modulation, which is used to produce the high-frequency pulsating DC voltage by means of applying six-pulse information on an average [13–16]. This will be done by the special coding in the PIC microcontroller.

1.2 Block Diagram of the Existing Figure 1 shows the block diagram of the existing system [20]. In this system, the fuel cell was connected to two isolated DC–DC converter and one three-phase inverter. The SRSPM technique will be used. Here, the single-reference six-pulse modulation technique is implemented to control the DC–DC converter. Also, a 12 V battery is energizing the auxiliary parts of the vehicle, since to stabilize the voltage variations of a fuel cell, a bidirectional DC–DC converter is used. The load is energized by the output of a three-phase inverter which is controlled by HF two-stage inverters [17–20].

Implementing PV Energized SRSPM-Based Single-Phase Inverter … 100V

32V – 68V

12 V Ba ery Bi-Direc onal

DC

DC

DC

Fuel Cell

233

DC

HF AC

DC

HF AC 6-Pulse Modula on

3Φ Inverter

IM

Pulsa ng DC

33% Modula on

Fig. 1 Block diagram of the existing system

1.3 Scope of Proposed System The one isolated DC–DC converter stage and one three-phase inverter stage are to be used in the proposed system instead of the two stages. In this system, we are replacing fuel cell into PV panel which is conventional source of energy. And also, we are using one isolated DC–DC converter instead of two stages to reduce the core loss. Figure 2 represents the block diagram of the proposed system, which have the following blocks; they are, • • • • • • •

PV panel Isolated DC–DC converter Three-phase inverter DC voltage sensor Gate driver Microcontroller (SRSPWM) Three-phase induction motor (vehicle motor).

Fig. 2 Block diagram of proposed system

234

P. Abirami et al.

The electric power from solar source is generated by using the PV panel. A photovoltaic module panel which is packaged and connected to assembly of the photovoltaic solar cell array is known as the PV panel. The output of the PV Panel is DC voltage. The DC voltage is applied to the DC–DC converters. Here, DC voltage is converted into AC voltage by using the power electronics device MOSFET, and the converted AC is applied to rectifier circuit which is also designed by the power electronics device MOSFET. The gate pulse of the each MOSFET is applied from the gate driver circuits. The gate circuits are controlled by the PIC microcontroller. These have the special hybrid technology SRSPM coding which working is given below. The proposed system can employ a single-stage rectifier/inverter topology. But high-frequency output needs a multistage inverter to boost its low voltage to generate three-phase voltage signals. High-frequency modulation technique is used to reduce the cost and weight of the system. Here, the six-pulse information is maintained at the DC link by activating the front-end full-bridge converter which in turn helps in obtaining balanced three-phase output voltage. In this research work, a singlereference modulation technique is proposed to maintain the voltage at the inverter terminals. At the front end, two bridge circuits are introduced to improve the power transferring capacity of the total unit. Though two bridges are connected, the system can be able to work with a single-bridge unit also. Both the front-end bridges are given an identical single reference. If any one of the bridges fails to work, the other bridge will be able to produce the six-pulse pulsating waveform and thus feeding the induction motor with the balanced three-phase voltage.

2 Simulation Results and Discussions 2.1 Open-Loop Control of the Proposed System Figure 3 shows the simulation circuit of the open-loop control of the proposed system. Here, we replaced the two sets of the DC–DC converters with two high frequency transformer with one set of the DC–DC converters with only one no. of highfrequency transformer to run the three-phase induction motor with single-reference six-pulse modulation. Figure 4 clearly shows the speed and torque of the open-loop controlled proposed system’s induction motor. Fig. 4 is drawn for the speed and torque of the induction motor with respect to the time which is in x-axis. In y-axis, it has the speed in RPM and torque in Nm. From the graph, we can observe the speed of the induction motor is 1200 RPM and its torque is 20 Nm. The voltage and current waveform of the induction motor in open-loop controlled proposed system is indicated in Fig. 5. From the graph, it is observed that the voltage and current of the induction motor are around 150 V and 20 A. Also, an unstable voltage and current are observed from the simulation results.

Implementing PV Energized SRSPM-Based Single-Phase Inverter …

Torque in Nm

Sped in RPM

Fig. 3 Open-loop control of proposed system

Time in Sec

Current in Amps

Voltage in volts

Fig. 4 Speed and torque of the induction motor

Time in Sec

Fig. 5 Voltage and current of the induction motor

235

236

P. Abirami et al.

Voltage in volts

Figure 6 displays the pulses of the semiconductor switches which are used in the converter circuit. In above graph, the converter pulse values are in volts and are displayed in y-axis with respect to the time in x-axis. Figure 7 displays pulses of the semiconductor switches used in the inverter circuit. In the above graph, the inverter pulse values are in volts and are displayed in y-axis with respect to the time in x-axis.

Time in Sec

Voltage in volts

Fig. 6 Pulse of the converter circuit

Time in Sec

Fig. 7 Pulse of the inverter circuit

Implementing PV Energized SRSPM-Based Single-Phase Inverter …

237

2.2 Closed-Loop Control of the Proposed System by PID Controller Figure 8 shows the simulation circuit of the closed-loop control of the proposed system by PID controller. Here, we replaced the two sets of the DC–DC converters with two high-frequency transformers with one set of the DC–DC converters with only one no. of high-frequency transformer to run the three-phase induction motor with single-reference six-pulse modulation. Here, the torque and speed values of the induction motor are measured and fed as a input to the PID controller and compared with a reference value to generate an error signal which is used to adjust the pulses of the inverter switches to obtain more accurate and desired value of output. Figure 9 clearly shows the speed and torque of the closed-loop controlled proposed system’s induction motor by PID controller. Figure 9 is drawn for the speed and torque of the induction motor with respect to the time which is in x-axis. In y-axis, it has the speed in RPM and torque in Nm. From the graph, it is observed that the speed of the induction motor is 1500 RPM and its torque is 4 Nm. Figure 10 indicates the voltage and current of the closed-loop controlled proposed system’s induction motor by PID controller. In Fig. 10, voltage and current of the induction motor are in x-axis with respect to the time which is in x-axis. From the

Torque in Nm

Speed in RPM

Fig. 8 Closed-loop control of proposed system using PID controller

Time in Sec

Fig. 9 Speed and torque of the induction motor

P. Abirami et al.

Current in Amps

Voltage in volts

238

Time in Sec

Fig. 10 Voltage and current of the induction motor

graph, the voltage and current of the induction motor are observed as 380 V and 5 A which are again not a stable one.

2.3 Closed-Loop Control of the Proposed System by ANFIS Controller Figure 11 shows the proposed system’s fuzzy logic controller editor. Here, the error and change in error signals are assigned as inputs to the system. The Sugeno (FIS type) neuro-logic is used to obtain the output voltage f (u).

2.4 Rule Base or Decision Making Framing of fuzzy rules includes linguistic variables to obtain the accurate output values. In the exact formation of rules for the designed fuzzy system, the performance of the converter gets improved. The control rules that initiate the relationship between the fuzzy output and the fuzzy inputs are derived from general knowledge of the system behavior. Here, seven linguistic variables are assigned thus 49 rules can be framed from these values. The combination of the linguistic variables is fine-tuned by trial and error method. Table 1 represents the control rules assigned to the proposed system’s ANFIS controller. Here, max–min inference method is used to obtain the control decision. AND operator is described by minimum function, and OR operator is described by

Implementing PV Energized SRSPM-Based Single-Phase Inverter …

239

Fig. 11 Fuzzy controller logic

Table 1 Fuzzy control rule table Inputs

Change in error NB

NM

NS

Z

PS

PM

PB

Error (e)

NB

NB

NB

NB

NM

NS

NS

Z

NM

NB

NM

NM

NM

NS

Z

PS

NS

NB

NM

NS

NS

Z

PS

PM

Z

NB

NM

NS

Z

PS

PM

PB

PS

NM

NS

Z

PS

PS

PM

PB

PM

NS

Z

PS

PM

PM

PM

PM

PB

Z

PS

PS

PM

PM

PB

PB

maximum function. Seven linguistic variables are utilized in the framing of rules. So, totally 49 rules can be framed from this. The simulation circuit of the closed-loop controlled proposed system by ANFIS controller is given in Fig. 12. The actual output of the induction motor’s speed and torque values is measured and fed as input to ANFIS controller by comparing it with a reference input, and output of controller is used to modify the pulses of the semiconductor switches of the inverter circuit to obtain more accurate and desired value of output voltage. The following graphs represent the simulation results of ANFIS controlled proposed system.

240

P. Abirami et al.

Fig. 12 Simulation circuit for ANFIS controller

Torque in Nm

Speed in RPM

Figure 13 clearly shows the speed and torque of the closed-loop controlled proposed system’s induction motor by ANFIS controller. Thus, from the above waveform, we can observe that the stable and accurate waveforms are produced by this ANFIS controller when compared with PID controller. Figure 13 is drawn for the speed and torque of the induction motor with respect to the time which is in x-axis. In y-axis, it has the speed in RPM and torque in Nm. The speed is maintained at 1500 rpm, and the torque is 4 Nm. Figure 14 indicates the voltage and current of the closed-loop controlled proposed system’s induction motor by ANFIS controller. From the voltage and current waveforms of induction motor, it is clearly explained that stable voltage and current values are obtained by ANFIS controller. In Fig. 14, voltage and current of the induction

Time in Sec

Fig. 13 Speed and torque of the induction motor

241

Current in Amps

Voltage in volts

Implementing PV Energized SRSPM-Based Single-Phase Inverter …

Time in sec

Fig. 14 Voltage and current of the induction motor

motor are in x-axis with respect to the time which is in x-axis. From the graph, we can know that the voltage of the induction motor is 420 V and its current is 6 A.

3 Hardware Diagram and Description 3.1 Block Diagram of the Hardware The block diagram of portable hardware kit is explained in detail in Fig. 15. In the hardware kit, we used single-phase inverter instead of the three-phase inverter. So, we used single-phase inverter to provide the necessary input supply to energize a single-phase induction motor. A PV panel is used to drive the power from the solar to energize the isolated DC–DC converter. The single six-pulse modulation technique is used to obtain the outputs from the inverter and converters. The gate drivers are used to provide the necessary pulses to the switches of the converter and inverter circuits.

3.2 Circuit Diagram of the Hardware Figure 16 shows the diagrammatical representation of the portable hardware kit’s circuit. In this circuit, we have the solar point which is used to receive the input supply from the solar panels. The solar panel input is applied to the DC–DC converter in

242

Fig. 15 Block diagram of hardware kit

Fig. 16 Circuit diagram of hardware

P. Abirami et al.

Implementing PV Energized SRSPM-Based Single-Phase Inverter … Table 2 Specification of induction motor

Table 3 Experimental output of the prototype kit

243

Parameters

Rating

Input voltage (V)

200–220 V

Input current (A)

2A

Power rating (HP)

0.33 HP

No. of phase

Single phase

Speed of the motor (RPM)

1440 RPM

S. No.

Parameters

Value

1

Input voltage (V)

30 V

2

Input current (A)

1A

3

Output voltage across the motor (V) 48 V

4

Output current (A)

0.3 A

5

Speed of the motor

180–190 RPM

between the high-frequency transformer which is located to change the frequency of the circuit to improve the output. The MOSFETs are used to convert the DC supply to AC Supply, then it is applied to the high-frequency transformer and updated highfrequency AC supply is applied to the diode rectifier circuit. The capacitor is used to filter the harmonics in the converted AC supply. That converted DC supply is fed to the inverter circuit. The inverted AC supply is applied to the single-phase induction motor. The voltage regulator circuit is used to provide the power supply to the MOSFET driver circuit. From the MOSFET driver circuit, the gate pulses are generated to activate the switches of inverter and converter circuits. The SRSPM coding is uploaded in the PIC microcontroller to generate the desired value of the gate pulse to MOSFET driver circuit. High output voltage is obtained to run single-phase induction motor. The specification of single-phase induction motor is given in Table 2. The experimental results like input voltage and input current of the inverter circuit and also output voltage and output current of induction motor are given in Table 3. It also tells about the speed of induction motor which is around 180–190 RPM.

3.3 Output Results of the Hardware Kit The input and output voltages of hardware arrangement are explained below by using CRO outputs. The input voltage readings of the batteries/PV panel are measured using CRO. Here, the input battery voltage is measured as 30 V. Figure 17 shows the output AC voltage readings of the induction motor which is measured from the CRO. From Fig. 17, the output AC voltage is measured as

244

P. Abirami et al.

Fig. 17 Output voltage

48 V. Figure 18 shows the input voltage given as input to the inverter circuit which is observed as 30 V. And also, Fig. 19 shows a tachometer which is used to measure the speed of the induction motor. From Fig. 19, the speed of the induction motor is measured as 187 rpm. Fig. 18 Input voltage

Implementing PV Energized SRSPM-Based Single-Phase Inverter …

245

Fig. 19 Speed of the motor = 187 RPM

4 Conclusion In this research work, the prototype of single-phase induction motor energized by solar energy was successfully implemented by using SRSPM-based single-phase inverter. To achieve compact, low-cost and lightweight system HF modulation is implemented in this system. Thus, to attain high-frequency output, a multistage inverter is required to boost the low voltage to generate the balanced three-phase voltage. Hybrid modulation comprises of pulsated DC voltage generation and 33% modulation to reduce the switching losses and switching frequency. The proposed work simulation is simulated under open-loop and closed-loop control. The closedloop control circuits have been simulated by applying the following two methods. I. PID controller II. ANFIS controller. From the simulation results, it is clear that the closed-loop system performs better than open-loop system and also ANFIS controller yields much better results than the PID controller. And also, solar energy is used to run the induction motor which is also toward sustainable low-carbon clean mobility owing to zero emission.

4.1 Future Scope As we have said, this research work can be implemented in renewable system (photovoltaic system) for the electrical vehicles. Here, the photovoltaic system is used as the resource. In future, the hybrid interpretation can be used for interconnecting the conventional energy resources to get better and continuous output to run the vehicle effectively.

246

P. Abirami et al.

References 1. Averberg A, Meyer KR, Mertens A (2008) Current-fed full-bridge converter for fuel cell systems. In: Proceedings of IEEE power energy society general meeting, pp 866–872 2. Aso S, Kizaki M, Nonobe Y (2007) Development of hybrid fuel cell vehicles in Toyota. In: Proceedings of IEEE power conversion conference, pp 1606–1611 3. Bilgin B, Emadi A, Krishnamurthy M (2010) Design considerations for a universal input battery charger circuit for PHEV applications. In: Proceedings of IEEE International Symposium on Industrial Electronics, pp 3407–3412 4. Emadi A, Williamson SS (2004) Fuel cell vehicles: opportunities and challenges. In: Proceedings of IEEE power energy society general meeting, pp 1640–1645 5. Emadi A, Williamson SS, Khaligh A (2006) Power electronics intensive solutions for advanced electric, hybrid electric, and fuel cell vehicular power systems. IEEE Trans Power Electron. 21(3):567–577 6. Emadi A, Williamson SS, Rajashekara K, Lukic SM (2005) Topological overview of hybrid electric and fuel cell vehicular power system architectures and configurations. IEEE Trans Veh Technol 54(3):763–770 7. Su G-J, Adams DJ, Peng FZ, Li H (2002) Experimental evaluation of a soft-switching DC/DC converter for fuel cell vehicle applications. In: Proceedings of IEEE Power Electronics and Transportation, pp 39–44 8. Huang R, Mazumder SK (2010) A soft switching scheme for multiphase dc/pulsating-dc converter for three-phase high-frequency-link pulse width modulation (PWM) inverter. IEEE TransPower Electron 25(7):1761–1744 9. Kim J-S, Ko J-M, Lee B-K, Lee H-B, Lee T-W, Shim J-S (2009) Optimal battery design of FCEV using a fuel cell dynamics model. In: Proceedings on Telecommunications Energy Conference, pp 1–4 10. Jang SJ, Won CY, Lee BK, Hur (2007) Fuel cell generation system with a new active clamping current-fed half-bridge converter. IEEE Trans Energy Convers 22(2):332–340 11. Jang SJ, Won CY, Lee BK, Hur J (2007) Fuel cell generation system with a new active clamping current-fed half-bridge converter. IEEE Trans Energy Convers 22(2):332–340 12. Khaligh A, Li Z (2010) Battery, ultra-capacitor, fuel cell, and hybrid energy storage systems for electric, hybrid electric, fuel cell, and plug-in hybrid electric vehicles: state of the art. IEEE Trans Veh Technol 59(6):2806–2814 13. Lesster LE (2000) Fuel cell power electronics—managing a variable-voltage dc source in a fixed-voltage ac world. Fuel Cells Bull 3(25):5–9 14. Mazumder SK, Rathore AK (2010) Performance evaluation of a new hybrid-modulation scheme for high-frequency-ac-link inverter: application for PV, wind, fuel-cell and DER/storage applications. In: Proceedings of IEEE Energy Conversion and Congr Expo, pp 2529–2534 15. Miller JM (2003) Power electronics in hybrid electric vehicle applications. In: Proceedings of 18th IEEE applied power electronics conference, vol 1. Miami Beach, FL, pp 23–29 16. Rajashekhara K (2003) Power conversion and control strategies for fuel cell vehicles. In: Proceedings of IEEE annual conference on IEEE industrial electronics society, pp 2865–2870 17. Vinod Kumar L, Nawaz SS (2016) Design and Implementation of Inverter for driving Induction Motor using DSPACE. Int J Res Appl Sci Eng Technol (IJRASET) 4(XI) 18. Schaltz E, Khaligh A, Rasmussen PO (2014) SRSPM methodology based on three phase inverter topology for PV system of interleaved high frequency. Int J Prof Eng Stud III(3) 19. Pal A (2018) A soft-switched high-frequency link single-stage three-phase inverter for grid integration of utility scale renewables. IEEE Trans Power Electron (99):1–1 20. Prasanna UR, Rathore AK (2013) A novel single reference six pulse modulation technique based interleaved high frequency three phase inverter for fuel cell vehicles. IEEE Trans Power Electron 28(12)

Enhancement of Voltage Transfer Ratio with Matrix Converter Based on Modulation Duty Cycle Matrix Approach Using Optimum AV Method N. Krishna Kumari, D. Ravi Kumar, and E. Shiva Prasad

Abstract Three-phase to three-phase matrix converters (MC) received significant interest in modern years since they are fine choice to inverters and cyclo converters. The reason is that MC offers (i) power flow in both the directions with sinusoidal input/output waveforms, (ii) better a better input power factor, (iii) compact design due to the nonexistence of DC-link capacitors which is used for energy storage purpose. In this chapter, two methods: Alesina–Venturini (AV) and optimum Alesina– Venturini (OAV) methods are considered using duty cycle matrix approach (DCMA). In the proposed modulation duty cycle strategy, the output voltage is obtained up to 48% with AV method and with OAV method 80% of the input voltage at different output frequencies. However, the input power factor is almost close to unity. The shapes of the output voltages output current and input current waveforms are observed. The proposed method is carried out using MATLAB/Simulink environment. The comparison is tabulated for both the methods. Keywords Alesina–Venturini method · Optimum Alesina–Venturini method matrix converter · Duty cycle matrix approach

1 Introduction The advancement of MC begins when Alesina and Venturini first anticipated the essential principles and operation in the early 1980s. The MC has several advantages over conventional rectifier-inverter type of power frequency converters: The absence of large reactive component made MC more advantageous over AC–AC converters and inverters, the reason is that the presence of reactive element deteriorates the N. K. Kumari · D. R. Kumar · E. S. Prasad (B) EEE Department, VNR VJIET, Hyderabad, Telangana, India e-mail: [email protected] N. K. Kumari e-mail: [email protected] D. R. Kumar e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_20

247

248

N. K. Kumari et al.

reliability of the system; it better inputs power factor; hence, close to sinusoidal input current and output current wave forms [1]; wide control of output voltage is possible with different output frequencies [2]. The main disadvantage is that it requires more number of semiconductor devices to provide bidirectional power flow, which increases the cost of the converter as well as complexity for generating signals. Consequently, discrete unidirectional devices are used for each bidirectional switch due to the nonavailability of monolithic bidirectional switches [3, 4]. On the other hand, there was a necessity of consistent bidirectional switches to obtain maximum voltage transfer ratio (VTR) and close to sinusoidal current waveforms with available modulation methods [5, 6]. In beginning, it was found that from the basic Alesina and Venturini’s method, the VTR was restricted to 0.5 with conventional pulse width modulation techniques. Later, it was improved to 0.866 by adopting third harmonic injection technique. This is an basic limitation in MCs even with balanced supply. Also SVM modulation technique gives the better voltage transfer ratio with optimum switching logic [2]. Patrick S. Flannery and Giri Venkataramanan proposed single antiparallel semiconductor quadrant devices to obtain bidirectional power flow with sinusoidal input and output current waveforms [7]. This arrangement establishes a path for the flow of current from inductive load to source, which avoids the voltage spikes during dead time [8]. Phase control of voltages in a power system by the means of a MC enables to shift voltage curves without changes of their amplitude [9].The direct MC follows the reference more accurately than the indirect MC [10]. Due to the presence of more number of switches in MC, these switches are subjected to less voltage stress. This feature enhances the reliability of the converter. This feature is compared and proved with a rectifier/inverter circuit for an aerospace motor drive application [11]. In some application where weight and volume are the main concern; for example in grid and transmission applications, isolated MC are preferred [12]. Different configurations of MCs are discussed with analysis in [13] and with SVM is in [14, 15]. However, single to three phase operation of MC is implemented in [16], and multiple operation converter is discussed in [17]. Vector control of a high performance drive using MC is executed in [18] and DTCMC converter for induction motor in [19, 20]. MC is a direct AC–AC converter, which replaces multiple conversion stages. It gives variable output voltage system with variable frequency. Circuit consists of an array of bidirectional switches arranged in such a way that any of the output lines of the converter can be connected to any of the input lines. This chapter introduces the topology of three-phase MC and analyzes the output of the converter for different output frequencies. The focus of the chapter is obtain output quantities at desired frequencies as well as to maintain sinusoidal shape of currents on input side as well as output side. The organization of the chapter is as follows: Sect. 2 describes the description of the modulation using duty cycle matrix approach (DCMA). Section 3 explains the modeling of the basic Alesina–Venturini method. Section 4 explains the modeling of the basic optimum Alesina–Venturini method (OAV). Section 5 discusses the comparative results obtained by the proposed AV and OAV with DCMA.

Enhancement of Voltage Transfer Ratio with Matrix Converter …

249

S 11

S 21

S 31

S 12

S 22

S 32

S 13

S 23

S 33

V i1 V i1 V i1

V01

V02

V03

Fig. 1 Basic circuit diagram of matrix converter

In this chapter, two methods, Alesina–Venturini (AV) method and optimum Alesina–Venturini (OAV) method, are proposed using DCMA. The basic circuit diagram of three-phase to three-phase MC is given in Fig. 1.

2 Modulation Using Duty-Cycle Matrix Approach (DCMA) In this chapter, modulation using DCMA is adopted to achieve the desired output waveforms at different frequencies. The relation between input voltages/currents and output voltages/currents is represented in a matrix form as given in Eqs. (1) and (2). ⎞ ⎛ ⎞⎛ ⎞ m 11 m 12 m 13 vi1 v01 ⎝ v02 ⎠ = ⎝ m 21 m 22 m 23 ⎠⎝ vi2 ⎠ v03 m 31 m 32 m 33 vi3 ⎞⎛ ⎞ ⎛ ⎞ ⎛ m 11 m 12 m 13 i 01 i i1 ⎝ i i2 ⎠ = ⎝ m 21 m 22 m 23 ⎠⎝ i 02 ⎠ i i3 m 31 m 32 m 33 i 03 ⎛

where, 0 ≤ m hk ≤ 1, h = 1, 2, 3, k = 1, 2, 3

(1)

(2)

(3)

The variables ‘mhk ’ are the duty-cycles of the nine switches which is represented in Fig. 1. From Fig. 1, 512 (29 ) possible switching states are possible taking into considerations of the following conditions; (a) The phases on the input side should not be short-circuited. (b) The phase currents of the output should not be open-circuited.

250

N. K. Kumari et al.

This implies that only one switch per output phase can be switched on at any instant of time. Considering all the above limitations, the total useful switching combinations are 27 in the circuit diagram of a three-phase to three-phase MC as shown in Fig. 1. To incorporate the necessary conditions which are listed in (a) and (b), the duty cycles of switches associated with three phase of a MC have to satisfy the constraints listed in Eqs. (4)–(6). m 11 + m 12 + m 13 = 1

(4)

m 21 + m 22 + m 23 = 1

(5)

m 31 + m 32 + m 33 = 1

(6)

Inductive load is considered at the output side to have continuous output current wave forms. However, capacitors are connected across each phase of a input supply to make sure continuous input voltage waveforms.

3 Alesina–Venturini Method Alesina–Venturini proposed the first solution with DCMA the control of input power factor, and output voltages with DCMA for unity power factor are obtained from Eq. (7) [1]. m hk =

1 2π 2π {1 + 2q(cos(α0 − (h − 1) )(cos(β1 − (k − 1) )} 3 3 3

(7)

where α0 βi Q H K

vector angle of the output voltage, vector angle of the input voltage, Vtr, Row value representation, and Column value representation.

The input voltage equations used in this chapter are given from Eqs. (8), (9) (Fig. 2). √ vi1 = 230 2 cos(ωi t) √ 2π vi2 = 230 2 cos(ωi t − ) 3

(8) (9)

Enhancement of Voltage Transfer Ratio with Matrix Converter …

251

Optimum AlesinaVenturini method

3-phase voltage Vi1 Source Vi2 Vi1 Vi2

Vi3

Vi3

Vi1

Ii1

Ii2 Ii3

Matrix converter circuit

Output Voltages and Currents Estimator for different output frequencies

I01 I02 I03

V01 V02 3-Phase load V03

g1,g2,…………….g9

Vi2 Vi3

Pulse Generator

Fig. 2 Implementation of the matrix converter OAV method

√ 2π ) vi3 = 230 2 cos(ωi t + 3

(10)

The input frequency ‘ωi ’ considered is 314.13 rads. Assuming that both the voltage vectors to be starting initially from zero degrees (reference) at t = 0, the angles are updated for every short time interval ‘t.’ Here, the output voltage vector angle is observed to be rotating ahead to input voltage vector, given its fraction of frequency. The angles are used at every time step t and substituted in Eq. (7) stated above. The modulation duty cycle matrix obtained is used for the calculation of instantaneous values of output voltages in Eq. (1) and input currents in Eq. (2). These instantaneous values at every time step t are used in plotting their respective waveforms as shown in the results below for the different values of the ‘q’ and the output frequencies. The initial work has been done for the output frequency with half of the input frequency. The voltage transfer ratio ‘q’ is taken for different values. In each case, the results taken are: (i) To derive power factor from input current and voltages (ii) Input/output current and voltages (iii) Total harmonic distortion analysis for output current and voltages.

4 Optimum Alesina–Venturini Method This is a modified method of Alesina–Venturini’s basic method of modulation. The major change in this modulation method is in the formula to calculate the modulation

252

N. K. Kumari et al.

duty cycle matrix ‘m,’ which enhance the voltage transfer capability of a MC. This modified formula is given in Eq. (11). ⎛

m hk

⎞ 1 + 2q(cos(β1 − (k − 1) 2π ) 3 1⎜ 2π 1 1 ⎟ = ⎝ (cos(α0 − (h − 1) 3 ) − 6 cos(3α0 ) + 2√3 cos(3βi ) ⎠ 3 2 2π 2π √ q(cos(4βi − (k − 1) ) − (cos(2βi + (k − 1) 3 3 3 3

(11)

Using this formula, the voltage transfer ratio has been improved to 87% with a reduced harmonic content in the waveforms of voltages and currents. As already seen that the output voltages can be obtained from the matrix relation among input parameters and the modulation duty cycle matrix ‘m,’ and the modulation duty cycle matrix ‘m’ has to be calculated using the equation for ‘mhk ’ for every time interval ‘t.’ So, it is evident that the angles of voltage vectors are to be obtained for each  time interval. Considering the output voltage vector angle α0 , let us define angle α0 as previous value of α0 and α a small incremental value to added to α0 . Then, α0 value can be obtained for every time interval t as shown below in Eq. (12). 

α0 = α0 + α

(12)

where α = ω0 t. The vector angle of input voltage βi can also be updated in the same procedure, wherein the equation is as given in Eq. (13). 

β0 = β0 + β

(13)

where β = ωi t. The updated angles from Eqs. (12) and (13) are substituted in the equation for modulation duty cycle matrix ‘m,’ i.e., Eqs. (7) and (11). To obtain the value of output voltage at t, the 3 × 3 duty cycle matrix is multiplied with the values of input voltages as given in Eqs. (8)–(10). The simple matrix representation can be given as: [v0 ]3×1 = [m]3×3 ∗ [vi ]3×1

(14)

From the output voltages obtained from Eq. (14), output currents can be found by applying KVL to the load side. The loads applied to a MC generally include R-L loads. The matrix representation of KVL at load side can be given as: [v0 ]3×1 = [R]3×3 ∗ [i 0 ]3×1 + [L]3×3 ∗ [di 0 ]3×1

(15)

where [R] and [L] are resistance and inductance matrices. The above equation can be represented in state space form as shown in Eq. (16) below.

Enhancement of Voltage Transfer Ratio with Matrix Converter … −1 [L]−1 3×3 [v0 ]3×1 − [i 0 ]3×1 [L]3×3 [R]3×3 ∗ [di 0 ]3×1

253

(16)

Solving the above ordinary differential equation gives the output current waveforms. The currents are also the plot of instantaneous values. These instantaneous values can now be used to trace the shape of input currents. Input currents are in relation to the output currents and the transpose of the modulation duty cycle matrix ‘m’ as explained in Eq. (17) can be given in simplified form as shown. T · [i 0 ]3×1 [i i ]3×1 = [m]3×3

(17)

Therefore, the waveforms of output voltages, output currents, and input currents can be plotted using the modulation duty cycle matrix ‘m’ which is being calculated at every time interval ‘t’ along with the instantaneous values at that time interval.

5 Simulation Results The work process was carried out in SIMULINK/MATLAB environment. Assuming balanced supply voltages and balanced output conditions, output frequency prefixed is half of the input frequency, i.e., ω0 = (ωi /2), ω0 = (ωi /3) as well as one third of the input frequency, i.e., ω0 = (ωi /4). The voltage transfer ratio ‘q’ has been varied from 0.33 to 0.48 in AV method and 0.6 to 0.8 in OAV method, and the waveforms of output phase voltage, output phase current, and input phase current have been plotted which are shown in Figs. 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, and 15. The

Fig. 3. Implementation of the MC with OAV method.

254

N. K. Kumari et al.

Fig. 4 Three-phase input voltage and current waveforms

Fig. 5 Output voltage, current, and input current waveforms for ω0 =

ωi 2

ω0 =

ωi 2

, q = 0.33 AVM

Enhancement of Voltage Transfer Ratio with Matrix Converter …

Fig. 6 THD analysis of output current waveform for ω0 =

ωi 2

ω0 =

Fig. 7 Output voltage, current, and current waveforms for ω0 =

ωi 3

255

ωi 2

, q = 0.33 AVM

ω0 =

ωi 3

, q = 0.33 AVM

256

Fig. 8 THD analysis of output current waveform for ω0 =

N. K. Kumari et al.

ωi 3

ω0 =

ωi 3

Fig. 9 Output voltage, current, and input current waveforms for ω0 =

, q = 0.33 using AVM

ωi 4

, q = 0.33 using AVM

Enhancement of Voltage Transfer Ratio with Matrix Converter …

Fig. 10 THD analysis of output current waveform for ω0 =

ωi 4

257

, q = 0.33 using AVM

Fig. 11 Output voltage, current, and input current waveforms for ω0 = OAVM

ωi 2

ω0 =

ωi 2

, q = 0.6 using

258

Fig. 12 THD analysis of output current waveform for ω0 =

N. K. Kumari et al.

ωi 2

ω0 =

ωi 3

Fig. 13 Output voltage, current, and input current waveforms for ω0 =

, q = 0.6 using OAVM

ωi 3

, q = 0.6 using OAVM

THD values of output phase voltage, output phase current, and input phase current are tabulated for each value of ‘q’ variation and with three output frequencies which are given in Tables 1 and 2 using AV and OAV with duty cycle matrix approach. From the THD values given in Tables 1 and 2, some inferences can be explained. The THD values of output voltages reduce as voltage transfer ratio (q) increases from

Enhancement of Voltage Transfer Ratio with Matrix Converter …

Fig. 14 THD analysis of output current waveform for ω0 =

ωi 3

259

, q = 0.6 using OAVM

Fig. 15 Output voltage, current, and input current waveforms for ω0 =

ωi 4

, q = 0.6 using OAVM

0.33 to 0.5 in AV method/0.6–0.8 in OAV method and remain constant with respect to output frequency variation. THD value of output current increases with respect to voltage transfer ratio and reduces for lower frequencies. The same explanation is valid for input current THD values.

260

N. K. Kumari et al.

Table 1 THD values for ωo = (ωi /2) using AV method in percentages Parameter

V0

I0

Ii

ω0 = ωi/2

ω0 = ω0i/3

ω0 = ω0i/4

VTR

VTR

VTR

0.33

0.4

0.4

0.48

0.33

0.4

0.48

V01

81

78.77

74.31

0.48

81.0

0.33

79.36

72.55

80.85

89.32

82.85

V02

81

78.77

74.31

81.0

79.36

72.55

80.85

92.44

86.46

V03

81

78.77

74.31

81.0

79.36

72.55

80.85

88.50

82.11

I01

2.43

2.37

2.16

2.2

1.95

1.81

2.29

1.85

1.76

I02

2.41

2.36

2.16

2.2

1.98

1.80

2.27

1.87

1.77

I03

2.42

2.35

2.18

2.2

1.98

1.81

2.28

1.88

1.75

Ii1

9.90

9.91

9.90

6.7

6.90

7.12

4.04

3.94

3.84

Ii2

9.50

9.38

9.26

5.4

5.22

5.08

4.16

4.10

4.04

Ii3

10.80

11.01

11.20

5.4

5.33

5.21

4.93

5.06

5.21

Table 2 THD values for ωo = (ωi /2) using OAV method in percentages ω0 = ωi/2

Parameter

ω0 = ω0i/3

VTR Vo

Io

Ii

ω0 = ω0i/4

VTR

VTR

0.60

0.70

0.80

0.60

0.70

0.80

0.60

0.70

0.80

V01

70.81

64.75

59.24

71.20

66.27

60.27

70.74

65.00

5924

V02

71.09

65.85

60.33

70.99

66.18

60.63

70.45

64.76

59.04

V03

70.58

64.72

59.34

71.16

66.22

60.78

71.03

65.30

59.51

I01

9.59

9.90

9.79

8.54

8.51

9.31

9.22

9.11

9.18

I02

9.81

9.29

9.34

854

8.51

9.27

9.34

9.16

9.19

I03

9.27

9.17

9.65

8.49

8.51

9.33

9.19

9.21

9.37

Ii1

22.76

20.38

19.09

9.58

8.58

7.50

9.93

8.97

8.15

Ii2

23.52

20.53

18.09

12.64

11.52

10.84

10.07

8.82

8.00

Ii3

23.46

20.62

18.92

19.51

17.17

1633

10.02

8.81

8.03

6 Conclusion The MC with DCMA using Alesina–Venturini (AV) method and Optimum Alesina– Venturini (OAV) method has been presented. In this chapter, the voltage transfer ratio ‘q’ has been varied from 0.32 to 0.48 with AV method and 0.6 to 0.8 with OAV method. Voltages and currents on source side as well as on load side are maintained sinusoidal in shape. The input power factor has been maintained at unity in all the cases. The THD values of the output currents and voltages are around acceptable values. MC can be required where a back-to-back voltage or current link systems are needed. The different applications of the converter include direct torque and fluxoriented control of motor drives, wind power generation in both full-power converter

Enhancement of Voltage Transfer Ratio with Matrix Converter …

261

topologies and doubly fed induction generator topologies. A method to enhance the ride through capability of the MC is by using the clamp capacitor to store the energy from the motor inertia. The compact nature of the converter is utilized to pack it inside an electric motor, and the overall system performance is successful. Recently, an increasing number of papers investigating the advantages/limitations of the use of MC in more electric aircrafts and battlefield tanks are also being reported. This might well be a potential application area of MC in the coming future.

References 1. Alesina A, Venturini MGB (1989) Analysis and design of optimum amplitude nine-switch direct AC–AC converters. IEEE Trans Power Electron 4:101–112 2. Casadei D, Serra G, Tani A, Zarri L (2002) Matrix converter modulation strategies: a new general approach based on space-vector representation of the switch state. IEEE Trans Ind Electron 49(2):370–381 3. Rao AK, Chatterjee JK, Subramanian S, Rajasekhar V (2010) Improved operation of a three phase matrix converter using simple modulation strategy. In: Joint international conference on power electronics, drives and energy systems (PEDES) & 2010 Power India 4. Simon O, Braun M (2001) Theory of vector modulation for matrix converters. In: Proceedings on EPE conference, Graz, Austria, Aug 27–29 5. Yue Z (2011) A study of a novel AC-DC-AC matrix converter with high voltage transfer ratio. In: International conference on business management and electronic information (BMEI), 2011, vol 4, pp 721–725 6. Sweety Jose P, Chandra Deepika N, Nisha SN (2011) Applications of single-phase matrix converter. In: Proceedings of ICETECT, pp 389–391 7. Flannery PS, Venkataramanan G A new poly phase matrix converter topology. In: 31st annual conference of IEEE industrial electronics society (IECON 2005), 6–10 Nov 2005, pp 1202– 1209 8. Idris Z, Zaliha S, Noor M, Hamzah MK (2005) Safe commutation strategy in single phase converter. In: International conference on power electronics and drives systems, (PEDS 2005), 28 Oct–01 Nov, pp 886–891 9. Sobczyk TJ, Sienko T Application of matrix converter as a voltage phase controller in power systems. In: International symposium on power electronics, electrical drives, automation and motion, (SPEEDAM 2006), 23–26 May 2006, pp 1322–1325 10. Jussila M, Eskola M, Tuusa H (2005) Analysis of non-idealities in direct and indirect matrix converters. In: European conference on power electronics and applications 11. Wheeler PW, Clare JC, de Lillo L, Aten M, Whitley C, Bradley KJ, Towers G (2005) A comparison of the reliability of a matrix converter and a controlled rectifier-inverter. In: European conference on power electronics and applications 12. Nasir U, Costabeber A, Wheeler P, Rivera M, Clare J (2019) A three-phase modular isolated matrix converter. IEEE Trans Power Electron 34(12):11760–11773 13. Maheswari KT, Bharanikumar R, Bhuvaneswari S (2019) A review on matrix converter topologies for adjustable speed drives. Int J Innov Technol Exploring Eng (IJITEE) 8(5):53–57. ISSN: 2278-3075 14. Asjad S, Dhawal RK, Ansari M, Ansari S (2018) Modeling of three phase matrix converter, performance and analysis. Int J Innov Res Comput Commun Eng 6(3):2487–2493 15. Hemakesavulu O, Subbaiah P (2018) Implementation of space vector PWM technique sin matrix converter control. Int J Res Appl Sci Eng Technol (IJRASET) 6(V):2038–2044 16. Chaudhari A, Thengane M, Wadiw S, Mude S (2018) Matrix converter: a novel topology for single to three phase conversion. Int J Eng Res Electric Electron Eng (IJEREEE) 4(3):43–46

262

N. K. Kumari et al.

17. Sriranjani R, Bharathi M, Chandana PNS (2017) Multi operation of SPWM based single phase matrix converter. ARPN J Eng Appl Sci 12(15):4451–4457 18. Mandal A, Nigam MK (2012) Modelling and simulation of matrix converter fed induction motor drive using PID control. Int J Eng Res Technol (IJERT) 1(10):1–7 19. Gaeid KS, Murshid AM, Salih ZH (2012) Direct torque control of induction motor with matrix converter. J Eng Sci Technol Rev 9(2):50–58 20. Rezaoui MM, Nezli L, Mahmoudi MO (2014) High performances of five-phase induction machine feeding by a matrix converter. J Electric Eng 65(2):83–89

Design and Implementation of Feedback Controller for Nonisolated Switching DC-DC Buck Converter Operating in Continuous Current Conduction Mode Manthan Mangesh Borage, Dipen M. Vachhani, and Rajesh Arya

Abstract DC-DC converters are one of the major research areas in the field of power electronics. Design-oriented study of lead compensator as well as Type-3 compensator is carried out and compared for its effectiveness in control of DC-DC buck converter. Design and development of laboratory prototype of 5 V, 3 A voltageregulated buck converter operating at switching frequency of 25 kHz is carried out involving design and fabrication of power circuit, MOSFET gate-driver circuit, and controller circuit. The control circuit is implemented using PWM controller IC SG3525AN. The bode diagrams of control to output transfer function uncompensated as well as compensated loop gain are plotted and verified to achieve required gain crossover frequency and phase margin with high loop gain at low frequencies leading to zero steady-state error. Developed laboratory prototype of the buck converter is thoroughly tested on resistive load and experimentally verified for its dynamic performance as well as stable close loop operation. Keywords Buck converter · Point of load converter · Lead compensator · Type-3 compensator · Gain crossover frequency · Control to output transfer function · Uncompensated system · Compensated system · Loop gain · Steady-state error

1 Introduction Buck converter is nonisolated type of switching DC-DC converter [1, 2]. It is the most common and simplest converter. It is also known as step-down converter as the output voltage (V ) is always less than the input voltage (V g ). Polarity of the output voltage is same as that of the input voltage. Figure 1 shows the basic circuit diagram of buck converter. Switch Q is semiconductor switch. D is the free-wheeling diode. M. M. Borage (B) SELECT, Vellore Institute of Technology, Vandalur-Kelambakkam Road, Chennai 600127, Tamil Nadu, India e-mail: [email protected] D. M. Vachhani · R. Arya Laser Electronics Division, Raja Ramanna Centre for Advanced Technology, Indore 452013, Madhya Pradesh, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_21

263

264

M. M. Borage et al.

Fig. 1 Circuit diagram of buck converter

Switch Q is repeatedly made on and off. This switching action causes a train of pulses which is filtered by second-order low-pass filter formed by inductor, L and capacitor, C to generate low ripple DC output voltage (V ). Buck converter as well as other nonisolated DC-DC switching converters is widely used as point-of-load (POL) power supply to provide well-regulated low value DC output voltage to variety of electronic loads from an intermediate DC input voltage bus or from battery voltage in battery powered appliances such as mobile phones and laptop computers where precise regulation of output voltage and fast dynamic response are major requirements apart from high-efficiency and small size [3]. Voltage-mode control and current-mode control are the commonly used PWM control techniques for control of DC-DC switching converters [2, 3]. Voltage-mode control of the buck converter with proportional controller is described in reference [4]. Compensator design procedure for buck converter with voltage-mode error amplifier is given in reference [5]. This paper reports design-oriented study, implementation as well as experimental testing and validation of voltage-mode controlled buck converter operating in continuous current conduction mode (CCM) with Type-3 compensator. Section 2 of the paper discusses modelling as well as compensator aspect related to closed loop control of DC-DC converter. Section 3 describes design, implementation, and development of laboratory prototype of voltage-mode controlled buck converter and presents experimental results obtained during testing of the converter verifying the design.

2 Closed Loop Control of DC-DC Converter Block diagram for closed loop voltage-mode control of buck converter is shown in Fig. 2 [2]. In voltage-regulated buck converter, output voltage, v(t), is required to be maintained constant against disturbances in input voltage, vg (t) and load current iload (t) as well as changes in circuit parameters. To achieve this, we introduce a compensator in the forward path whose main function is to adjust duty cycle, d(t) in such a way that desired output voltage is maintained regardless of line and load disturbances. For proper and stable closed loop operation, the compensator is to be designed in such a way that, it provides high gain at low frequencies, low gain at

Design and Implementation of Feedback Controller …

265

Fig. 2 Block diagram of buck converter operating in closed loop voltage-mode control

high frequencies, and adequate phase margin between 45° and 80° with required gain crossover frequency. Let, Gvd (s) Gvg (s) Z out (s) T (s) Gc (s) GPWM (s) H(s)

Control to output transfer function Line to output transfer function Output impedance Open loop transfer function or loop gain Transfer function of the compensator Transfer function of the PWM comparator Transfer function of sensor

Loop gain is given by, T (s) = G c (s)G PWM (s)G vd (s)H (s)

2.1 Lead Compensator Lead compensator is used to improve the phase margin of the system [2]. Transfer function of the lead compensator is,   G c (s) = G co (1 + s/ωz )/ 1 + s/ωp A zero is added to the loop gain, at a frequency f z , sufficiently far below the crossover frequency, f c such that the phase margin of the loop gain T (s) is increased by the required amount. Compensator also adds pole at high frequency which has beneficial effect of attenuating switching frequency noise. Disadvantage of lead compensator is that it gives low gain at low frequency which gives nonzero steady-state error in the converter output voltage. Therefore, the controller should be of Type-3 controller which has three poles and two zeros including one pole at origin.

266

M. M. Borage et al.

2.2 Type-3 Compensator Type-3 compensator provides three poles and two zeros including one pole at origin [5, 6]. Transfer function of Type 3 compensator is,     G C (s) = (G CO (1 + s/ωz )(1 + s/ωz1 ))/ s/ωz1 1 + s/ωp 1 + s/ωhp It is modified version of lead compensator with one additional pole at origin (f po ), one additional zero (f z1 ) sufficiently below crossover frequency (f c ) and one more additional pole (f hp ) sufficiently far away from crossover frequency (f c ). It is used when more than 90° of phase boost is necessary with high gain at low frequency.

3 Lab Prototype and Experimental Results In this section, design and implementation aspects of laboratory prototype of 5 V, 3 A closed loop voltage-mode controlled voltage-regulated buck converter operating at switching frequency of 25 kHz are described and obtained experimental results, and while testing, the developed converter is presented as well as discussed.

3.1 Buck Converter Design Specifications A buck converter and its compensator are designed and developed to comply with specifications as given in Table 1 Table 1 Design specifications S. No.

Specification/parameter

Value

1

Input voltage (V g )

15 V

2

Switching frequency (f s )

25 kHz

3

Load resistance (R)

1.667 

4

Peak-to-peak value of saw-tooth waveform (V m )

2.4 V

5

Output voltage (V )

5V

6

Filter inductor (L)

150 µH

7

Output voltage ripple (V )

≤50 mV

8

Reference voltage (V ref )

5V

9

Filter capacitor (C)

220 µF

10

Crossover frequency (f c = f s /10)

2.5 kHz

11

Phase margin (PM)

60°

12

Steady-state error

0

Design and Implementation of Feedback Controller …

267

Design Calculations For the CCM buck converter, Duty cycle, D = (V /V g ) = (5/15) = 33.33%, load current, I = (V /R) = (5/1.667) = 3 A. Here, H is DC gain of the output voltage sensor, which is given by I /H = V /Vref , Hence, H = 1. Transfer function of the PWM comparator is given by, G PWM (s) = 1/Vm Control to output transfer function for buck converter operating in CCM [2],   G vd (s) = Vg / 1 + Ls/R + s 2 LC For our designed buck converter, control to output transfer function is given by,    G vd (s) = 15/ 1 + 150 × 10−6 s/1.667 + s 2 150 × 10−6 × 220 × 10−6

(1)

Figure 3 shows bode plot of control to output transfer function. Normalized form of control to output transfer functions is,   G vd (s) = 15/ 1 + s/Qωo + s 2 /ωo2

Fig. 3 Bode plot of control to output transfer function

(2)

268

M. M. Borage et al.

Fig. 4 Bode plot of loop gain of uncompensated system

Comparing Eqs. (1) and (2) gives, quality factor, Q = 2.0188 and ωo = 5.504 krad/s. Angular crossover frequency, ωc = 2 × 3.14 × f c = 2 × 3.14 × 2.5 × 103 = 1.57 × 104 rad/s. We know for an uncompensated system, Gc (s) = 1. Uncompensated loop gain is given by,     T (s) = Vg × H /Vm × 1/ 1 + Ls/R + s 2 LC   T (s) = (15 × 1)/2.4 × / 1 + 150 × 10−6 s/1.667 + s 2 150 × 10−6 × 220 × 10−6 . Figure 4 shows the bode plot of uncompensated loop gain of the system. Since phase margin required for compensated system is required to be 60°, phase boost is required at gain crossover frequency. For uncompensated system, phase margin (PM) at gain crossover frequency, P M = 180◦ − 169◦ = 11◦ Thus, phase boost (θ ) of 60° − 11° = 49° is required. So, the Type-3 compensator will be designed according to these parameters. Here, f c = 2.5 kHz, θ = 49°, and gain of 1.34 dB or 1.668 at f c = 2.5 kHz. Using the design equations given in reference [2], Zero frequency,   f z = f c (1 − sin(θ ))/(1 + sin(θ ))0.5   f z = f c (1 − sin(49))/(1 + sin(49))0.5   f z = 2.5 × 103 (1 − sin(49))/(1 + sin(49))0.5 = 660.5285 Hz,

Design and Implementation of Feedback Controller …

269

ωz = 2 × 3.14 × f z = 4.1502 krad/s Pole frequency,   f p = f c (1 + sin(θ ))/(1 − sin(θ ))0.5   f p = f c (1 + sin(49))/(1 − sin(49))0.5   f p = 2.5 × 103 (1 + sin(49))/(1 − sin(49))0.5 = 9462.1 Hz, ωp = 2 × 3.14 × f p = 59.452 krad/s DC gain of the compensator, 0.5  G co = f z / f p × 10(1.34/20) = 0.3064 For removing switching frequency noises, we will put high frequency pole at 10 × f c . So, high frequency pole, f hp = 10 × f c = 25 kHz, ωhp = 2 × 3.14 × f hp = 1.57 × 105 rad/s. First zero frequency, f z1 = f c /10 = 250 Hz, ωz1 = 2 × 3.14 × f z1 = 1.57 krad/s. Pole at origin frequency, f po = Gco × f z1 = 76.6 Hz, ωpo = 2 × 3.14 × f po = 481.5 rad/s. Transfer function of Type-3 compensator,     G G (s) = (G CO (1 + s/ωz )(1 + s/ωz1 ))/ s/ωzl 1 + s/ωp 1 + s/ωbp       = G co ωp ωhp /ωz × ((ωz + s)(ωz1 + s))/ s ωp + s ωhp + s The loop gain of compensated system is, T (s) = G c (s)H (s)G PWM (s)G vd (s) Since the system is unity feedback, H(s) = 1, T (s) = G c (s)G PWM (s)G vd (s) Figure 5 shows bode plot of designed Type-3 compensator transfer function along with equivalent lead compensator transfer function. Figure 6 shows bode plot of loop gain of the compensated system with Type-3 compensator confirming phase margin of 60° at gain crossover frequency of 2.5 kHz.

3.2 Implementation of Type-3 Compensator Using Op Amp Using the following relations [6], we will find the values of required resistances and capacitance (Fig. 7).

270

Fig. 5 Bode plot of equivalent lead and Type-3 compensator

Fig. 6 Bode plot of loop gain of compensated system Fig. 7 Type-3 compensator using op amp

M. M. Borage et al.

Design and Implementation of Feedback Controller …

271

Let R1 = 100 k     C1 = f hp − f z / 2π R1 f po f hp C 1 = 20.2 nF (selected value (10 nF + 10 nF) = 20 nF) C2 = ( f p − f z1 )/(2π R1 f p f z1 ) C2 = 6.2 nF (selected value (3.3 nF + 3.3 nF) = 6.6 nF) C3 = f z /(2π R1 f po f hp ) C3 = 540 pF (selected value 470 pF) R2 = (R1 f po f hp )/(( f p − f z ) f z ) R2 = 11.9 k (selected value 12 k) R3 = (R1 f z1 )/( f p − f z1 ) R3 = 2.71 k (selected value 2.7 k)

3.3 Circuit Diagrams The detailed circuit diagram used for implementation of power board, control board, and driver board of the laboratory prototypes of close loop voltage-mode controlled buck converter operating at switching frequency of 25 kHz are shown in Fig. 8. The list of their key components is given in Tables 2, 3 and 4, respectively. Because of the floating source of the buck converter MOSFET switch, isolated gate driver is required, which is implemented using optocoupler-based gate driver IC FOD3180 [7] and isolated supply on the MOSFET side is obtained by IC DCP021515P [8]. Controller circuit is implemented using PWM controller IC SG3525AN [9] as shown in the diagram. Figure 9 shows the developed laboratory prototype of the buck converter with all subsystems.

3.4 Experimental Results The developed prototype of closed loop controlled buck converter is thoroughly tested on resistive load for steady-state as well as dynamic performance, and following experimental waveforms are obtained. Figure 10 shows key waveforms of the buck

272

M. M. Borage et al.

Fig. 8 Detailed circuit diagram of developed laboratory prototype of power board (top), control board (bottom right), and driver circuit (bottom left) Table 2 Components list for power board S. No.

Component description

Quantity

1

MOSFET switch (STP20NF06L)

2-Nos.

2

Inductor (150 µH–31 turns on toroidal core—HF1061252)

1-No.

3

Schottky diode (SB540)

1-No.

4

Al polymer capacitor (220 µF, 16 V)

1-No.

5

Polypropylene capacitor (1 µF, 63 V)

2-Nos.

6

Al polymer capacitor (10 µF, 25 V)

1-No.

Table 3 Components list for driver board S. No.

Component description

Quantity

1

Optocoupler gate driver IC (FOD3180)

1-No.

2

Isolated DC-DC converter IC (DCP021515P)

1-No.

3

Fast recovery diode (1N914)

1-No.

4

Ceramic capacitor (0.1 µF, 63 V)

3-Nos.

5

Polyester capacitor (1 µF, 63 V)

2-Nos.

6

Electrolytic capacitor (10 µF, 25 V)

2-Nos.

Design and Implementation of Feedback Controller … Table 4 Components list for control board

S. No.

Component description

Quantity

1

PWM controller IC (SG3525AN)

1-No.

2

Fast recovery diode (1N914)

2-Nos.

3

Polyester capacitor (3.3 nF, 63 V)

2-Nos.

4

Polyester capacitor (10 nF, 63 V)

2-Nos.

5

Polyester capacitor (6.8 nF, 63 V)

1-No.

6

Polyester capacitor (470 pF, 63 V)

1-No.

7

Electrolytic capacitor (10 µF, 25 V)

2-Nos.

Fig. 9 Laboratory prototype of the buck converter Fig. 10 Waveforms (from top) of inductor current, MOSFET gate-source voltage, MOSFET drain-source voltage, and output voltage

273

274

M. M. Borage et al.

converter operating in steady-state continuous current conduction mode while delivering load current of 3 A at regulated output voltage of 5 V. Load transient test is also performed on the converter to verify closed loop stable operation as well as dynamic performance. Figure 12 shows the waveforms of output voltage and inductor current during 1 A of step change in load current. Figure 11 shows magnified view of output voltage undershoot when step load of 1 A is applied on the converter. Figure 13 depicts magnified view of output voltage overshoot when step load of 1 A is removed. In both cases, following the transient, output voltage reaches final value promptly without oscillations, verifying proper controller design and stable operation of control loop of the buck converter. Fig. 11 Waveforms of output voltage (top), inductor current (middle), and load current (bottom) when 1 A of step load is applied

Fig. 12 Load transient waveforms for 1 A step change in load current: output voltage (top), inductor current (middle), load current (bottom)

Design and Implementation of Feedback Controller …

275

Fig. 13 Waveforms of output voltage (top), inductor current (middle), load current (bottom) when 1 A of step load is removed

4 Conclusion Design-oriented study of buck converter compensator design was carried out for their operation under continuous current conduction mode in closed loop voltagemode control. Lead compensator was attempted first, but it suffered from nonzero steady-state error though it provided required crossover frequency and phase margin. Therefore, a Type-3 controller was designed and implemented using op amp present in the PWM controller IC SG3525AN. The optocoupler gate driver IC FOD 3180 is used to drive the MOSFET switch. The bode plots of control to output transfer function and loop gain (compensated and uncompensated) were plotted and verified for achievement of required phase margin and gain crossover frequency. Buck converter was implemented on hardware in the laboratory. Testing of the developed converter was carried out, and proper closed loop stable operation of the converter was verified with satisfactory dynamic as well as steady-state performance.

References 1. Hart DW (2011) Power electronics. McGrow Hill (India) Private Limited 2. Erickson RW, Maksimovic D (2000) Fundamentals of power electronics, 2nd edn. Springer Science Publication 3. Maksimovic D, Zane R, Erickson R (2004) Impact of digital control in power electronics. In: Proceedings of the 16th international symposium on power semiconductor devices and ICs— 2004 4. Kazimierczuk MK, Sathappan N, Czarkowski D (1993) Voltage-mode-controlled PWM buck DC-DC converter with a proportional controller. In: Proceedings of the IEEE 1993 national aerospace and electronics conference—NAECON 5. Rahimi AM, Parto P, Asadi P. Compensator design procedure for buck converter with voltagemode error-amplifier. Application Note AN-1162, International Rectifier

276

M. M. Borage et al.

6. Demystifying type II and type III compensators using op-amp and OTA for DC/DC converters; Application report, SLVA662—July 2014, Texas Instruments 7. Datasheet of gate drive optocoupler, FOD3180. Available online at following url: https://www. onsemi.com/pub/Collateral/FOD3180-D.pdf 8. Datasheet of isolated DC-DC converter, DCP021515P. Available online at following url: http:// www.ti.com/lit/ds/symlink/dcp022405.pdf 9. Datasheet of PWM controller IC, SG3525AN. Available online at following url: https://www. onsemi.com/pub/Collateral/SG3525A-D.pdf

Design and Analysis of Improved Indirect Matrix Converter Supplying Power to Rotor of DFIG for Bi-directional Power Flow N. Lavanya and P. N. H. Phanindra Kumar

Abstract This paper proposes an improved switching strategy to control both input and output stages of a three-phase Indirect Matrix Converter (IMC). In improved switching, the input stage uses largest positive phase to phase voltage at each instant to produce maximum DC voltage with low ripple, and hence, the switching losses and the IMC output voltage distortions are reduced compared to Conventional Indirect Matrix Converter. Also, the fundamental input current of IMC with improved control is at unity power factor (UPF) for all operating conditions. The main advantage of IMC with constant input voltage and frequency is to supply power to multiple loads at different voltages and frequencies at the output stages. In this paper, the performance of improved IMC with two output stages, where output stage-2 is supplying power to static load at 45 Hz and output stage-1 is fed to rotor of doubly fed induction machine (DFIM) for bi-directional power flow is analyzed. The hardware setup for the proposed IMC is implemented in the laboratory and the experimental results are compared with those from simulation analysis. Keywords Improved switching scheme · Bi-directional power flow · UPF

1 Introduction Due to the high reliability and less maintenance cost of AC motors, variable voltage and variable frequency (VVVF) AC motor drives are gaining much popularity in commercial and industrial applications. The accessibility of excellent fast forced commutation devices and the development in digital computation technology have motivated a rapid growth of power converter, which further enhances the effective use of modem AC motor drives. The conventional AC/DC/AC power converter topology [1–3] is widely used in AC drives due to simple control and less cost. However, the N. Lavanya (B) Nalla Malla Reddy Engineering College, Hyderabad, India e-mail: [email protected] P. N. H. Phanindra Kumar Selec Controls Pvt. Ltd., Navi Mumbai, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_22

277

278

N. Lavanya and P. N. H. Phanindra Kumar

use of a energy storage element (i.e., electrolytic capacitor) in the DC-link makes the converter bulky and with limited lifetime is a major drawback of this AC-AC converter, which has lead to more research in alternative topologies one of it being the matrix converter (MC) [4, 5]. MC is a direct AC/AC power conversion topology, without the need of a bulky and short lifetime energy storage elements in the DC-link. Although MC has disadvantages such as low voltage transfer ratio (0.866), and the number power semiconductor devices required is more, still the MC has attained high research interest due to the following advantages such as regenerative power flow, high quality input and output waveforms, variable power factor and absence of bulky short lifetime energy storage elements at the DC-link. There are two types of MCs classified: Direct Matrix Converter (DMC) and Indirect Matrix Converter (IMC) [6]. The DMC consists of nine bidirectional switches, which are complex to control. But the IMC is a two-stage converter, comprises of four-quadrant voltage source rectifier connected to a two-level voltage source inverter. In some applications, IMC is preferred to DMC, due to its simpler and safe switching control; further, reduction of power semiconductor devices is possible in IMC and also the IMC is used to supply power to multiple loads. The IMC topology supplying to multiple loads is shown in Fig. 1. The main disadvantage of conventional IMC [7–10] is high ripples in DC output voltage which results in more switching losses at both the stages of IMC, distortions

Fig. 1 Proposed IMC topology

Design and Analysis of Improved Indirect Matrix Converter Supplying …

279

in IMC output voltage and also commutation problems at input stage. To overcome all these limitations, IMC discussed in this paper is with improved control at the input stage. In this paper, the performance of the proposed IMC [11–17] is analyzed for IMC supplying power to rotor of DFIM for speed control. The paper is organized into following sections. Section 2 deals with the improved switching technique used to control both the stages of proposed IMC. The performance of proposed IMC supplying power to rotor of DFIM is discussed in Sect. 3. The conclusions drawn in the paper are given in Sect. 4.

2 Proposed IMC 2.1 Input Stage The input stage of proposed IMC consists of six bi-directional switches as shown in Fig. 1. The three input supply phase voltages are given by Eq. (1). va = Vm sin ωt   2π vb = Vm sin ωt − 3   2π vc = Vm sin ωt + 3

(1)

where Vm is the maximum value of the input supply voltage and ω is its angular frequency. Six sectors are considered for input stage, each of 60° duration and their corresponding angles are shown in Table 1. Table 1 Sector-wise DC voltages at Input stage of conventional IMC θin

Sector

Largest phase--phase voltage

Second-largest phase--phase voltage

DC-link Voltage in a sampling period

0–60°

1

V cb

V ab

TPR · Vcb + TnR · Vab

60°–120°

2

V ab

V ac

TPR · Vab + TnR · Vac

120°–180°

3

V ac

V bc

TPR · Vac + TnR · Vbc

180°–240°

4

V bc

V ba

TPR · Vbc + TnR · Vba

240°–300°

5

V ba

V ca

TPR · Vba + TnR · Vca

300°–360°

6

V ca

V cb

TPR · Vca + TnR · Vcb

280

N. Lavanya and P. N. H. Phanindra Kumar

Fig. 2 Hardware setup of proposed IMC

3 For Conventional IMC The input phase voltage angle θ in (=ωt) which ranges from (0 to 360°) is divided into six sectors of 60° duration as shown in Fig. 2. For any sector, the maximum DC-link voltage is obtained by using largest and second-largest phase to phase voltages in a given sampling period. Detailed analysis is given in papers [11–17]. Table 1 gives the sector-wise DC voltages generated at the input stage of conventional IMC. The fundamental input current of conventional IMC is at UPF, but the DC output voltage will have high ripples, which causes distortion in output voltage of IMC which is explained in following sections. To overcome this, deficiency IMC with improved control is proposed.

4 Improved IMC In this proposed switching strategy, maximum DC voltage is obtained by using largest positive phase to phase input voltage in each sector. In sector-1, the maximum voltage Vcb is obtained at the DC terminals by turning ON the switches Scp and Sbn . The DC-link voltages for the remaining sectors are shown in Table 2. This proposed switching strategy gives maximum DC voltage with low ripple at the output of the rectifier stage compared to that of conventional IMC, where the DC voltage is obtained by using first-largest and second-largest phase to phase input voltages in each sector.

Design and Analysis of Improved Indirect Matrix Converter Supplying … Table 2 Sector-wise DC voltages at input stage

Identification of different sectors 1 2 3 4 5 6

θ in −π π 6 to 6 π π 6 to 2 π 5π 2 to 6 5π 7π 6 to 6 7π 3π 6 to 2 3π −π 2 to 6

281

  DC-link voltage Vpn (Vc − Vb ) = Vcb Vab Vac Vbc Vba Vca

4.1 Output Stage The inverter stage of proposed IMC is controlled using space vector pulse width modulation explained in [11–17]. Since the DC voltage of IMC is not constant but it is variable, the improved switching times for the output stage are given by Eqs. (2) and (3) for conventional IMC and proposed IMC as shown below. For conventional IMC, the switching times are T A = T A

T pR Ts

, TB = TB

T pR Ts

, T A = T A

TnR  TR , TB = TB n and TZ = Ts − (T A + TB ) Ts Ts (2)

where TpR and TnR are the switching times for the upper(p-group) devices and the lower (n-group) devices used in the input stage. Similarly, the new switching times for the proposed IMC are given by T A =

TA  TB  T A  TB , TB = , TA = , TB = , and TZ = Ts − (T A + TB ) 2 2 2 2

(3)

Similarly for other sectors, the switching times are calculated for the output stages of both the IMCs. It can be observed that the switching scheme for the IMC with proposed control is simpler and independent of input stage control unlike the switching scheme for the conventional IMC. Table 3 gives the switching sequences for sectors 1–6 as shown. In the proposed IMC, this new switching sequence at the inverter stage is used to minimize output voltage distortion and also to reduce the switching losses. The hardware setup for proposed IMC is divided into three parts: power circuit, signal conditioning unit and controller as shown in Fig. 2.

282

N. Lavanya and P. N. H. Phanindra Kumar

Table 3 Switching sequence for inverter stage of proposed IMC

Sector number

Switching sequence

1

V1 − V2 − (111)VZ − V2 − V1

2

V2 − V3 − (000)VZ − V3 − V2

3

V3 − V4 − (111)VZ − V4 − V3

4

V4 − V5 − (000)VZ − V5 − V4

5

V5 − V6 − (111)VZ − V6 − V5

6

V6 − V1 − (000)VZ − V1 − V6

5 Performance of IMC Supplying to Rotor of DFIM and Static Load The circuit topology for this application is shown in Fig. 3, where output stage-1 is connected to rotor of DFIM and output stage-2 supplies power to static load at 45 Hz. The performance of proposed IMC for bi-directional power flow between output stage-1 and rotor of DFIM is analyzed. The rotor speed of the DFIM is given by Eq. (4), Nr =

120 ∗ ( f s − f r ) P

(4)

where f s and f r are the frequencies of applied voltage to the stator and the rotor. The power flow in DFIM is analyzed by using per phase steady-state equivalent circuit as shown in Fig. 4.

3 phase AC supply

Rectifier

Inverter1

IM

Inverter2

RL Load

3 phase AC supply

IMC

Fig. 3 Proposed IMC supplying DFIM and static load

Fig. 4 Equivalent circuit of DFIM

Rs

jX ls

Ir

jX lr Rr s

Vs

jX m AC

Vr s

Design and Analysis of Improved Indirect Matrix Converter Supplying …

283

where V s and V r are stator and injected rotor voltages referred to stator, respectively. The Air gap power, Pg = 3 Ir Torque, T = 3 Ir

(Ir Rr + Vr ) s

(5)

(Ir Rr + Vr ) ωs s

(6)

The power flow in DFIM is controlled by varying the output voltage and frequency of the IMC through change of reference voltage V R (square root of inverter output phase voltage) of the output stage and its frequency (ωr ). The experimental analysis is carried out by varying the output stage-1 voltage and frequency for power flow in DFIM for sub-synchronous speeds at 1200 and 900 rpm and output stage-2 reference voltage and frequency is set to 100 V and 45 Hz, respectively. Here, the stator voltage and frequency of DFIM are set to 220 V (lineline) and 50 Hz. For a given rotor speed (N r ) of DFIM, the output stage frequency is r . set to f r = f s − P∗N 120 DFIM operating at 1200 rpm: As output stage reference voltage V R increases, the developed electromagnetic torque in DFIM gradually changes. Figures 5 and 6 are obtained through simulation as can be seen the rotor feeds power into the IMC when V R is 10 V at frequency of 10 Hz. The fundamental input voltage and current of proposed IMC are at UPF. Figures 7 and 8 are the experimental results obtained for this case. Figures 9, 10, 11 and 12 show the simulation and experimental results when the power flow of DFIM is from IMC to the rotor when the reference voltage is 30 V and frequency is 10 Hz.

Voltage(V)

DFIM operating at 900 rpm speed: Figs. 13 and 14 show the simulation and experimental results of DFIM where the power flow is from rotor to IMC. Here, the output stage of IMC reference voltage is 35 V and frequency is set to 20 Hz. Figures 15 and 16 show the simulation and experimental results when the power

200 0 -200 0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

0.92

0.94

0.96

0.98

1

Time(sec)

Current(A)

5

0

-5 0.8

0.82

0.84

0.86

0.88

0.9

Time(sec)

Fig. 5 Input voltage and current

Voltage(V)

284

N. Lavanya and P. N. H. Phanindra Kumar 20 0 -20 0.7

0.75

0.8

0.85

0.9

0.95

1

0.85

0.9

0.95

1

Time(sec) Current(A)

5

0

-5

0.7

0.75

0.8

Time(sec)

Fig. 6 Output stage-1 (line-line) voltage and current Input voltage and current

200 V/div 10 ms/ div

Fig. 7 Input voltage and current of IMC Output line to line voltage

Output current A

50V/div 10A/div 50ms/div

Fig. 8 Output stage-1 phase voltage and current Voltage(V)

200 100 0 -100 -200

0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

0.92

0.94

0.96

0.98

1

Time(sec) Current(A)

2 1 0 -1 -2

0.8

0.82

0.84

0.86

0.9

0.88

Time(sec)

Fig. 9 Input voltage and current of proposed IMC

Design and Analysis of Improved Indirect Matrix Converter Supplying …

285

Voltage(V)

100 50 0 -50 -100 0.7

0.75

0.8

0.7

0.75

0.8

Time(sec)

0.85

0.9

0.95

1

0.85

0.9

0.95

1

Current(A)

1 0.5 0 -0.5 -1

Time(sec)

Fig. 10 Output stage-1 (line-line) voltage and current Input voltage and current

200 V/div 10ms/ div

Fig. 11 Input voltage and current of proposed IMC Inverter - 1 output voltage

output current

50 V/div, 1A/div, 10ms/div

Fig. 12 Output stage-1 (line-line)voltage and current

Voltage(V)

100 50 0 -50 -100 0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

0.92

0.94

0.96

0.98

1

Time(sec)

Current(A)

10 5 0 -5 -10 0.8

0.82

0.84

0.86

0.88

0.9

Time(sec)

Fig. 13 Input voltage and current of proposed IMC

286

N. Lavanya and P. N. H. Phanindra Kumar Input phase voltage and current

200 V/div 10 A/div 10ms/div

Fig. 14 Input voltage and current

Voltage(V)

100 50 0 -50 -100

0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

0.92

0.94

0.96

0.98

1

Time(sec)

Current(A)

2 1 0 -1 -2

0.8

0.82

0.84

0.86

0.88

0.9

Time(sec)

Fig. 15 Input voltage and current for proposed IMC

Input phase voltage and current

200 V/div 2 A/div 10ms/ div

Fig. 16 Input voltage and current

flow in DFIM is from IMC to rotor at output reference voltage V R = 75 V and frequency 20 Hz. Here, the fundamental input voltage and current of proposed IMC are at UPF.

6 Conclusion The IMC with improved switching scheme is used to control both input and output stages is proposed. This improved switching strategy gives maximum DC voltage with low ripple at the output of the rectifier stage compared to that of conventional

Design and Analysis of Improved Indirect Matrix Converter Supplying …

287

IMC. Also, the IMC is used to supply power to rotor of DFIM for bi-directional power flow from rotor to IMC and vice-versa. The hardware setup for the proposed IMC is implemented in the laboratory and it is tested for different operating conditions. Both the experimental and simulation results are presented in the paper and also the fundamental input voltage and current of proposed IMC are at UPF.

References 1. Pena R, Clare JC, Asher GM (1996) Doubly fed induction generator using back-to-back PWM converters and its application to variable speed wind-energy generation. IEE Proc Electr Power Appl 143(3) 2. Friedli JT, Kolar J (2010) Comprehensive comparison of three-phase AC–AC matrix converter and voltage dc-link back-to-back converter systems. In Proceedings of IPEC, pp 2789–2798 3. Cruz SMA, Ferreira M (2009) Comparison between back-to-back and matrix converter drives under faulty conditions. In: Proceedings of 13th European Conference on Power Electronics and Application, Barcelona, Spain, pp 1–10 4. Wheeler PW, Rodríguez J, Clare JC, Empringham L, Weinstein A (2002) Matrix converters: a technology review. IEEE Trans Industr Electron 49(2):276–288 5. Ghoni R, Abdalla AN, Koh SP, Rashag HF, Razali R (2011) Issues of matrix converters: technical review. Int J Phys Sci 6(15):3576–3588 6. Pena R, Cardenas R, Reyes E, Clare J, Wheeler P (2009) A topology for multiple generation system with doubly fed induction machines and indirect matrix converter. IEEE Trans Ind Electr 56(10):4181–4193 7. Nguyen TD, Lee H-H (2014) A new SVM method for an indirect matrix converter with common-mode voltage reduction. IEEE Trans Ind Inf 10(1) 8. Pena R, Cardenas R, Reyes E, Clare J, Wheeler P (2011) Control of a doubly fed induction generator via an indirect matrix converter with changing DC voltage. IEEE Trans Ind Electron 58(10):4664–4674 9. Nguyen T, Lee H-H (2012) Modulation strategies to reduce common mode voltage for indirect matrix converters. IEEE Trans Ind Electron 59(1):129–140 10. Chen X, Kazerani M (2006) A new direct torque control strategy for induction machine based on indirect matrix converter. In: IEEE international symposium on industrial electronics, vol 3, pp 2479–2484 11. Lavanya N, Venu Gopala Rao M (2015) Control of indirect matrix converter with improved SVM method. Int J Power Electr Drive Syst 6(2):370–375 12. Lavanya N, Chandra Sekhar O, Ramamoorty M (2016) Performance of modified indirect matrix converter with multiple loads at different frequencies. Tech J Inst Eng (India) 40: 220–225 13. Lavanya N, Chandra Sekhar O, Ramamoorty M (2016) Performance of indirect matrix converter as asynchronous link between two Ac systems. J Electr Eng 16(49):434–442 14. Lavanya N, Chandra Sekhar O, Ramamoorty M, Venu Gopala Rao M (2017) Performance of modified indirect matrix converter supplying power to static load. Int J Control Theor Appl 10(5): 367–376 15. Lavanya N, Chandra Sekhar O, Ramamoorty M (2017) Performance of indirect matrix converter with improved control feeding passive and induction motor. J Adv Res Dyn Control Syst 9(6): 243–263 16. Lavanya N, Chandra Sekhar O, Ramamoorty M (2017) Performance of indirect matrix converter with improved control feeding doubly fed induction machine. IEEE SPICES 8–10:1–6 17. yLavanya N, Chandra Sekhar O, Ramamoorty M (2017) Performance of indirect matrix converter with improved control feeding to induction motor for speed control by using pi and fuzzy controllers In: IEEE TENCON, 5–8 Nov 2017

Dynamic Analysis of Fuel Cell-Fed Superlift Converter R. Hounandan, V. Chamundeeswari, M. S. Kumaran, and J. Deepa

Abstract As renewable energy plays a vital role in today’s world, this paper presents a non-pollutant renewable source namely the fuel cell-fed DC-DC converter. Fuel cell is an electrochemical cell that produces electrical energy from a chemical reaction. Here, a superlift DC-DC converter is interfaced with a fuel cell and simulated. As the output voltage is solely dependent on the fuel cell voltage, the characteristics of the fuel cell is predicted by carrying out dynamic analysis. It reveals the mathematical model of fuel cell which portrays its behavioral nature. Different types of fuel cells are considered, but proton exchange membrane (PEM) fuel cell is chosen because of its efficient characteristics. Simulation is done using MATLAB, and the results are validated with the theoretical calculations. Keywords PEM fuel cell · Dynamic model · Superlift converter

1 Introduction Nowadays, the industries are expecting renewable energy sources such as fuel cells which produce electric current without any emission of pollutant gases. There are many types of fuel cells of which the proton exchange membrane (PEM) fuel cell [1] is the most popular and more efficient. In order to condition the output of a fuel cell, a DC-DC [2] converter is interfaced with the PEM fuel cell which converts DC voltage level into the desired required magnitude. With this, the analysis of the PEM cell with a superlift converter is carried out. The superlift converter chosen here for analysis is improved, and the negative output superlift luo converter (INOSLC) [3] is considered for analysis. Mathematical modeling is a technique used to formulate a model whose solution gives exact solutions and guidelines for the proper utilization R. Hounandan (B) · M. S. Kumaran · J. Deepa UG Scholar, Department of EEE, St. Joseph’s College of Engineering, Chennai 600119, India e-mail: [email protected] V. Chamundeeswari Associate Professor, Department of EEE, St. Joseph’s College of Engineering, Chennai 600119, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_23

289

290

R. Hounandan et al.

Fig. 1 Block diagram of the proposed system

of the system. The modeling of the power electronic converter [4] is carried out, and the equations are represented and the parameters are tabulated. A dynamic model represents the behavior of an object over time. It is used where the object’s behavior is best described as a set of states that occur in a defined sequence. In this work, a complete analysis of the circuit parameters and its feasibility with renewable energy is portrayed through mathematical modeling and dynamic analysis. The superlift luo converter [5] is fed to an R-Load, and its results are validated with theoretical calculations (Fig. 1).

2 Overview of the Proposed System Proton exchange membrane (PEM) fuel cell [6] consists of electrolyte which is acidic-based polymer membrane and platinum-based electrodes. It has the ability to satisfy dynamic power requirement. The basic block diagram of the fuel cell [7]-fed DC-DC converter is shown in Fig. 2. The fuel cell is a stack of many fuel cells Fig. 2 PEM fuel cell

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

291

combined. Each stack is capable of producing 0.7–1 V. Hydrogen is given as fuel to PEM [7] at the anode side. The bipolar layer goes through a series of reactions when hydrogen incidents at anode. The other side at the cathode oxygen is supplied. 2H2 → 4H+ + 4e

(1)

O2 + 4H+ + 4e → 2H2 O

(2)

2H2 + O2 → 2H2 O

(3)

Equation (1) gives the reaction of hydrogen with the anode. The electrons released are tap between anode and cathode which result in the flow of electrons which is electric current. Now, the H+ ions move across the electrolytic membrane. Equation (2) gives the reaction of oxygen with the cathode. Now, the H+ ions combine with the O2 and e to produce water which escapes the fuel cell. The overall reaction is given as Eq. (3). Electric current, heat, and water are the output of these reaction, where the water finds its way out of the fuel cell via water exhaust.

3 DC-DC Superlift Converters and Its Mathematical Analysis The fuel cell is coupled with a DC-DC [8] converter to get the required DC output. The converter employed here is superlift converters. The conventional converter is negative output superlift luo converter (NOSLC) and the proposed is improved NOSLC (INOSLC). The explanation of the conventional is carried out followed by the proposed.

3.1 Mathematical Modeling of NOSLC As the negative voltages play a vital role in telecom, rig lines, and various other applications, it is preferred here. The elementary circuit of NOSLC [9] is taken. It consists of two diodes, one inductor and two capacitors. It also produces an enhanced output voltage with a gain value of 3. For example, if an input voltage of 12 V is given, it produces an output voltage of −36 V. Figure 3 represents the circuit of NOSLC. Figures 4 and 5 represent the circuit of NOSLC in ON and OFF mode. During ON mode, the inductor current L 1 rises and the capacitor voltage C 1 forward biases the diode D1 . Capacitor C 2 voltage is the load voltage. During OFF

292

R. Hounandan et al.

Fig. 3 Circuit of NOSLC

Fig. 4 Circuit of NOSLC–ON mode

Fig. 5 Circuit of NOSLC–OFF mode

mode, diode D2 forward biases and the voltage across the voltage will be equal to the capacitor C 1 and C 2 voltages. The equations during ON and OFF are explained. During mode 1, when the switch is ON, applying KVL to the loop, Vs = L

di dt

(4)

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

293

Fig. 6 Input voltage (Vin ), pulse (67%), and output voltage (Vo ) of NOSLC

Fig. 7 Improved INOSLC circuit

Fig. 8 Circuit of INOSLC in ON mode

=L

I t1

(5)

294

R. Hounandan et al.

Fig. 9 Circuit of INOSLC in OFF mode

Assuming the inductor current linearly from I 1 to I 2 in a time t 1 . Let t 1 = KT is the ON time of the switch. The peak-to-peak value of the current is given by, I =

Vs KT L

(6)

During mode 2 when is switch is OFF, the inductor current falls linearly from I 2 to I 1 in a time t 2 . Applying KVL to the loop 2, the voltage equation is given by, di dt

(7)

−LI t2

(8)

Vs − Vo = −L Vs − Vo =

Let t 2 = (1 − K)T is the OFF time of the switch. I =

(Vo − Vs )(1 − k)T L

(9)

where I is the peak-to-peak ripple current of inductor. Equating Eqs. (5) and (7), (Vo − Vs )t2 Vs t1 = L L

(10)

Vs (t1 + t2 ) = Vo t2

(11)

The voltage gain in terms of ON time and OFF time of the switch is given by, t1 + t2 Vo = Vs t1

(12)

Duty cycle (K) is defined as the ratio of ON time to the total time. K =

t1 t1 + t2

(13)

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

295

K (t1 + t2 ) = t1

(14)

K t2 = t1 (1 − K )

(15)

From Eq. (12), the gain in terms of the duty. The output voltage of the converter is given by, Vo =

Vs 1−K

(16)

The voltage gain is given by, Vo 1 = Vs 1−K During mode 1, the capacitor supplies the load current. The average capacitor current is equal to the load current at time t 1 . The ripple voltage on capacitor is given by, 1 Vc = C

t1 I cdt

(17)

0

Vc =

1 I a t1 C

(18)

Vc =

1 I a t1 C

(19)

Let IL be the average inductor current. Ripple current, I = 2IL

(20)

Critical value of the inductance is given by, Lc =

(1 − K ) R 2f

(21)

Let Vc be the average capacitor voltage. Ripple voltage, Vc = 2V a Critical value of the capacitance,

(22)

296

R. Hounandan et al.

Cc =

K 2fR

(23)

whereK is the duty ratio in percentage, R is the load resistance in Ohms, and f is the switching frequency in Hertz. Assuming duty ratio = 67%, switching frequency = 100 kHz, load resistance = 50  and by substituting these values in Eqs. (21) and (23), the values of L and C are determined. C1 = C2 = 30 μF L = 10 mH

(24)

3.2 Simulation Results of NOSLC The circuit is simulated with the given modeled parameters, and the simulation results are obtained (Fig. 6). For an input of 12 V, an output voltage of −36 V is obtained which is shown in the figure with a 67% duty cycle.

3.3 Mathematical Modeling of Improved NOSLC The proposed circuit is INOSLC [10]. The mathematical modeling of the improved NOSLC circuit is discussed in this section. The switching topologies of the proposed circuit are as follows. Figure 7 represents the circuit of improved INOSLC circuit. Figures 8 and 9 represent the circuit of INOSLC in ON and OFF mode. In this mode, the switch is closed and the source voltage results in flow of current through the inductor L1 and capacitor C1 . Since capacitor C1 has zero impedance to current, the capacitor C1 charges faster than inductor thus diode D1 gets forward biased. Equations during ON state are given as follows. For inductor current, I L 1 =

vin dt vin K T = L1 L1

I c20N = I0 =

CdVo dt

(25) (26)

In this mode, the switch is opened, thereby the energy that is stored in the inductor L 1 , L 2 and the capacitor C 1 discharges across the nodal points of the capacitor C 2 thus boosts the output voltage. The inductor L 1 supplies the load current.

Dynamic Analysis of Fuel Cell-Fed Superlift Converter Table 1 Parameters of INOSLC

297

Parameter’s name

Symbol Value

Input voltage

V in

20 V

Output voltage

Vo

121.21 V

Inductors

L1 , L2

0.01 mH, 0.01 mH

Capacitors

C 1, C 2

30 µF, 10 µF

Nominal switching frequency

Fs

50 KHz

Load resistance

R

100 

Duty cycle

K

0.67

Time period

T

0.02 ms

Variation ratio of output voltage 11

1.1 × 10–3

Variation ratio of output voltage 12

3.3 × 10–3

Equations during OFF,  v0 − vin IL = (1 − k)T L1   v0 − vin IL = (1 − k)T L2 

1

(27)

During steady state, the current through C2−ON time capacitor current equals C2−OFF time capacitor current. kTi C2−ON = (1 − k)Ti C2−OFF

(28)

Variation ratio of inductor current i L 1 is,  δ=

 R k(1 − k) G 2 f L1

(29)

Variation ratio of output voltage V o is (Table 1): δ=

(1 − k) 2 f RC2

(30)

3.4 Simulation Results of INOSLC The INOSLC with load resistance 100 , when fed with 20 V DC supply and duty ratio of 67% gives the output voltage of -123 V. The INOSLC is simulated in MATLAB, and the simulated outputs are shown in the following figures. Figure 10

298

R. Hounandan et al.

Fig. 10 Input voltage to the converter

50

Voltage(V)

40 30 20 10 0

0.02

0

0.04

0.06

0.08

0.1

Time(s)

shows the input voltage of 20 V. The voltage across the capacitor C 1 is 17 V as shown in Fig. 11. Figure 12 depicts the load voltage or the capacitor C 2 voltage as −121 V and the output current of the converter is depicted in Fig. 13 as −1.25 A. Fig. 11 Voltage across the capacitor C1

25

Voltage(V)

20 15 10 5 0

0.02

0

0.04

0.06

0.08

0.1

0.08

0.1

Time(s)

Fig. 12 Voltage across the capacitor C2

50

Voltage(V)

0 -50 -100 -150 -200

0

0.02

0.04

0.06

Time(s)

Dynamic Analysis of Fuel Cell-Fed Superlift Converter Fig. 13 Output current of the converter

299

0.5

Current(A)

0 -0.5 -1 -1.5 -2

0

0.02

0.04

0.06

0.08

0.1

Time(s)

Fig. 14 Fuel cell-fed improved NOSLC converter

4 Dynamic Analysis of PEM Fuel Cell Fuel cell depends upon various operational parameters such as relative humidity, temperature, pressure, membrane thickness, anodic and cathodic stoichiometric flow, and distribution of oxygen. The most critical problems to overcome in the PEM fuel cells [11] are the thermal and water management and the variation of the fuel cell voltage due to the partial pressures of H2 and O2 . The dynamic analysis of the problems is discussed as following.

4.1 Thermal and Water Management The moisture content in the polymer electrolyte membrane and its temperature are very important for the reactivity of the gases on the electrodes. So, proper humidification has to be provided to the membrane. If in case excess water is present in

300

R. Hounandan et al.

the membrane sites, it leads to the reduced reactivity of the gases on the electrodes decreasing the fuel cell output. And increase in the cell temperature affects the electrochemical reaction rates drastically [6]. Thus, both the moisture content and the temperature must be properly handled for better performance of the fuel cell [8, 9]. Water flux of a membrane is defined as the volume per unit area per unit time. It defined the rate of the water in a membrane. To manage the moisture in the membrane, the water flux is obtained by Nernst--Plank equation.   1 1 − k = (0.005139λ − 0.00326) exp 1268 303 T 

(31)

Also, the water diffusion coefficient in the membrane is expressed as, ⎧ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨

⎫ ⎪ 1forλ ≤ 2 ⎪ ⎪ ⎪ ⎪ 1 + 2(λ − 3)for2 < λ ≤ 3 ⎬ DT = 3 − 1.38(λ − 3)for3 < λ ≤ 4 ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ 2.563 − 0.33λ2 ⎪ ⎪ ⎪ ⎩ +0.0264λ2 − 0.000671λ3 forλ > 4 ⎪ ⎭

(32)

where DT is expressed as, Dw = Dλ

(33)

   1 1 − DT = 106 exp 2416 303 T

(34)

Now, the boundary conditions are expressed as z = z c . λ| A−Cat = λ|mem

(35)

ϕ| A−Cat = ϕ|mem

(36)

N W |mem = cte = N W | A−cat

(37)

Equation (37) gives the water flux in the polymer electrolyte membrane. From this, the moisture in the membrane as well as the temperature of the fuel cell is optimized and the performance of the fuel cell is improved.

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

301

4.2 Variation of Fuel Cell Voltage The fuels such as hydrogen and oxygen are fed to the fuel cell, and as a result, the electrochemical reaction inside fuel cell results in electron flow which produces output voltage. The output voltage depends on the mixture of hydrogen (H2 ) and oxygen (O2 ) fed to the fuel cell. Thus, the proportion (i.e.) the partial pressures of H2 and O2 contributes to the output voltage [12]. The output voltage is derived from the Nernst equation. The proportional relationship of the flow of gas through a value with its partial pressure can be expressed as, kan qH2 = = kH2 pH2 MH2

(38)

q H2 O kan = = k H2 O pH2 O M H2 O

(39)

There are three relevant contribution to the hydrogen molar flow. They are, the input flow, the flow that takes part in the reaction, and the output flow. The relationship among the factors can be written as, d RT in PH2 = (q − qHout2 − qHr 2 ) dt Van H2

(40)

The basic electrochemical relationship between hydrogen flow and the stack current can be written as, qHr 2 =

NI = 2kr IFC 2F

(41)

Using (40) and (41) and applying Laplace transform, the hydrogen partial pressure can be written as, PH2 =

 1/k H2  in q H2 − 2kr I FC 1 + τ H2 S

(42)

where τH2 =

Van kH2 RT

(43)

Similarly using (42), the partial pressure of water PH2 O and oxygen O2 can be determined. As the performance of the fuel cell is affected by the flow of gases to the electrodes, the fuel cell output voltage maybe expressed as, Vcell = E + ηact + ηohmic

(44)

302

R. Hounandan et al.

Let us consider the system having constant temperature and concentration, Vcell = E − Bln(C I ) − R int I FC ηact + ηohmic

(45)

Therefore, the Nernst voltage in terms of gas molarities can be written as, E = N [E 0 +

pH pO 0.5 RT log( 2 2 )] 2F pH2 O

(46)

Equation (46) gives the relationship between the partial pressures of hydrogen, oxygen and water H2 O, and the output voltage. From this equation, the partial pressures of gases can be properly managed and the maximum voltage can be achieved [13].

5 Interfacing of INOSLC with FUEL CELL The interfacing of PEM fuel cell with INOSLC [14] is implemented, and the simulation results are shown. Table 2 shows the parameters of INOSLC with fuel cell. The simulation are done in MATLAB with the above Fig. 14 and the results are shown below. Figure 15 shows the PEMFC [15] of rating 50 kw 625 V DC interfaced with INOSLC and simulated and the respective voltage and current waveforms are depicted. It produces an output of −1800 V and −18 A current. Figure 16 shows the PEMFC of rating 50kw 625 V DC interfaced with INOSLC and simulated and the respective voltage and current waveforms are depicted. It produces an output of −400 V and −2.5 A current. Figure 17 shows the PEMFC of rating 50kw 625 V DC interfaced with INOSLC and simulated and the respective voltage and current waveforms are depicted. It produces an output of −190 V and −1.9 A current. Figure 18 shows the AFC 2.4 kW 48 V DC interfaced with INOSLC and simulated and the respective voltage and current waveforms are depicted. It produces an output of −430 V and −2.7 A current. Table 2 Parameters---INOSLC with fuel cell Type of fuel cell

Fuel cell wattage (kilowatt)

Fuel cell voltage (Volts)

Output voltage (Volts)

Output current (Amperes)

PEM fuel cell

1.26

24

−190

−1.9

PEM fuel cell

6

45

−400

−2.5

PEM fuel cell

50

652

−1800

−18

Alkaline fuel cell

2.4

48

−430

−2.5

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

Fig. 15 Waveforms of PEMFC of rating 50kw 625 V

Fig. 16 Waveforms of PEMFC of rating 6 kW 45 V DC

Fig. 17 Waveforms of PEMFC of rating 1.26 kW 24 V DC

303

304

R. Hounandan et al.

Fig. 18 Waveforms of AFC 2.4 kW 48 V DC

6 Conclusion A fuel cell-based improved luo converter is taken for analysis in this work and simulated, and the results are validated with the theoretical calculations. PEM fuel cell of different wattages are taken and analyzed. Modeling of the improved luo converter is carried out and the circuitry parameters are found out. Dynamic analysis of the fuel cell is done based on the parameters like heat and water management and variation of fuel cell voltage, and the characteristics is inferred with the derived equations.

References 1. Lai J-S, Ellis MW (2017) Fuel cell power systems and applications. In: Proceedings of the IEEE, vol 105, no 1 2. Fang Lin Luo (2014) Hong Ye, Super-lift boost Converters, IET Power Electr, Vol 7. Iss 7:1655–1664 3. Cocor A, Baescu A, Florescu A, Stoichesu D-A (2015) Elementary and self-lift negative output luo Dc-Dc converters used in hybrid cars. U.P.B. Sci Bull Ser 77(4): 179–190 4. Zhu M, Wang T, Luo FL (2012) Analysis of voltage-lift-type boost converters, 7th IEEE conference on ICIEA 5. Luo FL (2000) Double output Luo converters, an advanced voltage technique. IEEE Power Electr 147(6):469–484 6. Guaitolini SVM, Yahyaoui I, Fardin JF, Encarnaço LF (2018) A review of fuel cell and energy cogeneration technologies. In: 2018 9th international renewable energy congress (IREC)

Dynamic Analysis of Fuel Cell-Fed Superlift Converter

305

7. Shahin A, Hinaje M, Martin J-P, Pierfederici S, Raël S, Davat B (2010) High voltage ratio DC–DC converterfor fuel-cell. In: IEEE transactions on industrial electronics, vol 57, no 12 8. Jiao Y, Luo FL, Zhu M (2011) Voltage lift type switched inductor cells for enhancing DC-DC boost ability: principles and integrations in Luo converter. IET Power Electr 4(1):131–141 9. Luo FL, Ye H (2003) Positive output super lift converters. IEEE Power Electr 18(1):105–113 10. Luo FL (2008) Analysis of super lift Luo converters with capacitor voltage drop. IEEE 18(6):417–422 11. Smith JD, Novy M (2018) Design of a modern proton-exchange membrane fuel cell module for engineering education. IEEE, 2018 IEEE conference on technologies for sustainability (SusTech) 12. He Y, Chen H, Quab B, Zhang T Pei P, Liang C (2018) Analysis of proton exchange membrane fuel cell reactant gas dynamic response and distribution quality. Energy Proc 152:667–672 13. Cultura AB,Salameh ZM (2014)Dynamic analysis of a standalone operation of PEM. Fuel Cell Syst. doi: 10.4236/jpee.2014.21001 14. Kirubakaran A, Jain S,Nema RK (2009) The PEM fuel cell system with DC/DC boost converter: design, modeling and simulation. Int J Recent Trends Eng 1(3) 15. Liu H, Song Q, Zhang C, Chen J, Deng Bo, Li J (2019) Development of bi-directional DC/DC converter for fuel cell hybrid vehicle. J Renew Sustain Energy 11:044303. https://doi.org/10. 1063/1.5094512.p

DSPACE1103 Controller for PWM Control of Power Electronic Converters R. Amalrajan, R. Gunabalan, and Nilanjan Tewari

Abstract This paper presents the overview and working procedure of dSPACE1103 controller and its applications for power electronics converters. The configuration and control desk parameter settings have been discussed in detail. The pulse width modulation (PWM) for duty cycle control of DC-DC converter and sinusoidal pulse width modulation (SPWM) for voltage and frequency control of DC-AC converter using real-time interface (RTI) library blocks in dSPACE and MATLAB simulink blocks in MATLAB simulink environment are presented. The simple blocksets in simulink and RTI library are used to obtain the control signals in order to turn on and turn off the power semiconductor switches in converters. The programming skills and knowledge are not required to generate the control signals. The real-time prototype implementation of DC-DC converter and DC-AC converter in MATLAB with dSPACE controller is presented. Experimental results are provided to enhance the performance of dSPACE controller. Keywords dSPACE1103 controller · DC-DC converter · DC-AC converter · Pulse width modulation (PWM) · Sinusoidal pulse width modulation (SPWM)

1 Introduction In the present scenario, the applications of DC-DC converter are increased in medical equipment [1], DC distribution [2–4], and standalone renewable energy systems [5]. Similarly DC-AC converter usage along with DC-DC converter is increased in hybrid electric vehicle [6–8], smart grid [9–11], solid-state transformers [12, 13] and R. Amalrajan (B) · R. Gunabalan · N. Tewari School of Electrical Engineering, Vellore Institute of Technology, Chennai 600127, Tamil Nadu, India e-mail: [email protected] R. Gunabalan e-mail: [email protected] N. Tewari e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_24

307

308

R. Amalrajan et al.

inverter-based distributed generation system [14–16]. As photovoltaics, batteries, and other renewable/alternative power sources continue their rapid growth, many inverter/converter topologies have been introduced in literatures [17]. The presence of more number of semiconductor switches reduces the efficiency and reliability of power electronic converter/inverter [18]. A unified PWM concept was introduced in SPWM, space vector PWM (SVPWM) [19], and discontinuous space vector PWM (DSVPWM) and applied to a three-phase dual-buck inverter to reduce the computational burden when implemented by a digital signal processor (DSP) [20]. A new simple and easy to use low cost, flexible, high accuracy digital photovoltaic array simulator with buck converter was designed [21]. The nonlinearity of I–V curve causes a higher order polynomial equation that increases the computational time, size of memory, and cost of the DSP. A buck-boost DC-DC converter with PV emulator using multiple simple I-V polynomial equations was implemented in low cost 8-bit microcontroller to overcome computational difficulties [22]. To implement P, PI, PID, and optimization techniques, control boards including field programmable gate array (FPGA), microcontroller and LabVIEW controller have been used. But these controllers could not be interfaced with MATLAB simulink during implementation. The regenerative braking and friction brake controller were integrated into the advanced vehicle simulator (ADVISOR) of electric vehicle (EV), and its performance was validated in real-time by LabVIEW real-time controller and hardware in the loop (HIL) test bench [23]. For efficiency improvement, these models can be implemented in dSPACE controller. DSPACE is widely used for planning, designing, execution of hardware and software, and to improvise the accuracy of the output. DSPACE is preferred mainly in mechatronic systems such as converter, inverter, multilevel inverter, and voltage and frequency (VFD) control for drive systems. This helps for development and testing activities in real time. The dSPACE controller is broadly applied in industries for controlling electric-powered automobile and power electronics converters [24]. It is highly useful for engineers as it gives high accuracy while working with real-time controllers, low power use, brief development period, and user-friendliness. The fuel cell (FC) model was implemented in MATLAB simulink and converted into a C program which was programmed into a dSPACE or eZdsp R2818 DSP controller [25]. The dSPACE controller was used for prototype design & testing, and DSP controller was used for field testing. A new lightning search algorithm was loaded in dSPACE DS1104 controller board for low power applications [26]. A solar emulator with better dynamics was built for different irradiations using DS1104 controller [27]. The robustness of the controller was demonstrated with 20% load disturbance and 10% input disturbance in a solidstate transformer using dSPACE 1104 controller and Opal-RT simulator [28]. The dSPACE1104 controller is a lowest version, and researchers are using the higher version of dSPACE1103 controller. The HIL setup was projected to test the adaptive headlight concept using dSPACE1103 in xPC target box platform [29]. The vehicle state information was sent by digital to analog channels of DS1103 and received by analog to digital channels of xPC target box. The dSPACE1103 controller was used for DC motor speed control applications [30].

DSPACE1103 Controller for PWM Control of Power Electronic Converters

309

In above literatures [24–30], dSPACE was used to generate PWM/SPWM in power electronics converters. The configuration parameter settings and design procedure are required for better understanding. In this paper, a different approach to generate control signals for power electronic converters in MATLAB simulink with dSPACE1103 controller is presented with configuration settings of MATLAB and control desk. The simulink block output is interfaced with real-time DC-DC converter and three-phase inverter using dSPACE1103 controller. The experimental setup and results of the converters are presented.

2 Methodology MATLAB and control desk have basic set of guidelines and few preliminary procedures. The major RTI library tools such as MASTER PCC and SLAVE DSP F240 are used to generate the control signal and interface the feedback signals. The RTI blocksets are used for pulse generation for DC-DC converter and DC-AC converter with simple simulink blocks. The model is built to obtain the control signal through dSPACE1103 controller output pins without delay in real time. The simple block diagram of interfacing dSPACE controller with sensors and converters is shown in Fig. 1. The control desk is a software tool which is interfaced with MATLAB and RTI library. The RTI library is a software tool which is added to MATLAB library in order to generate control signal and to build MATLAB simulink model to get the simulation results through dSPACE1103 controller. Similarly, the signals from the sensors are feedback through dSPACE1103 to the MATLAB simulink blocks. Both control signal and feedback signal can be controlled and monitored in control desk. The dSPACE controller is used to transfer & interface control signal and feedback signal to both hardware and software. The dSPACE controller is connected with desktop and experimental setup which is controlled through control desk/MATLAB simulink. The accessories of dSPACE1103 as shown in Fig. 2 consist of a connector and LED panels, controller board, and license key. The connector and

Fig. 1 Block diagram

310

R. Amalrajan et al.

Fig. 2 DSPACE1103 accessories

Fig. 3 Real-time interface (RTI) library

LED panels have easy access to I/O signals, ADC inputs, DAC outputs, digital I/O, slave DSP I/O, incremental encoder interfaces, and serial interfaces. The RTI library in Fig. 3 has master PPC and slave DSP. The slave DSP simulink in-built blocks are used to generate pulses. The slave DSP, master PPC, and MATLAB library simulink in-built blocks are used to develop a model for single-phase pulse generation, threephase pulse generation, and SPWM. The blocksets which are available in SLAVE DSP F240 and the corresponding output pin connection details are shown in Fig. 4. The master PPC simulink in-built blocks which are used to sense feedback signals and the corresponding output pin details in the digital I/O connector are indicated in Fig. 5.

DSPACE1103 Controller for PWM Control of Power Electronic Converters

311

Fig. 4 Slave DSP F240 and pin details of slave I/O connector

Fig. 5 Master PPC and pin details of digital I/O connector

3 Design Steps in MATLAB and Control Desk In MATLAB simulink model, the following steps to be followed for configuring parameters and interfacing the control desk:

3.1 MATLAB 1. 2. 3. 4. 5. 6.

Launch MATLAB version 2015a Create a new simulink model Design and save the model Set the stop time(inf) Open and edit the configuration parameters Set the solver parameters a. Solver_ode1(euler) b. Fixed step size multiples of sample time

312

R. Amalrajan et al.

c. Uncheck optimization 7.

Set the hardware configuration to be a. b. c. d. e. f.

Device vendor (Generic) Device type (Custom) Integer (int) Floating point (float) Byte ordering (BigEndien) Signed integer division rounds to (zero)

8. 9.

Click apply & ok Build the model and verify the steps in diagnostics viewer. The build process should be 100% successful 10. An.sdf file will be generated upon the completion of build process. On successful generation of.sdf file, the results will be validated in dSPACE ctrl_desk.

3.2 SPACE Control Desk 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Check the pulse results in the slave I/O connector MATLAB simulation block can be found in “all variable descriptions” Select.sdf file and choose model root Drag and drop MATLAB blocks to dSPACE home screen Choose appropriate measurement indicators Launch control desk on desktop Create a new project and experiment Choose names for project, root directory, and experiment Add platform device DS1103 from the list Import.sdf file (it will be available in MATLAB file save location) After completion of above steps go online

4 4. PWM Generation The PWM signal for DC-DC converter is generated using simple simulink blocks of RTI library. Figure 6 indicates the single-phase PWM generation using DS1103SL_DSP_PWM block in RTI library. DS1103SL_DSP_PWM, DS1103SL_DSP_PWM1, DS1103SL_DSP_PWM2, and DS1103SL_DSP_PWM3 blocks are used to generate constant duty cycle pulses for semiconductor switches 1 and 2. In DS1103SL_DSP_PWM and DS1103SL_DSP_PWM1 blocks, dead band period is not possible. In DS1103SL_DSP_PWM2 and DS1103SL_DSP_PWM3 blocks, user can set dead band limit up to 0–100 μs.

DSPACE1103 Controller for PWM Control of Power Electronic Converters

313

Fig. 6 PWM generation for DC-DC Converter

The following steps are followed for PWM pulse generation: Step 1 Step 2 Step 3 Step 4

go to new simulink model go to RTI library select the user requirment PWM simulink blocks set the initialization limits

The PWM block (DS1103SL_DSP_PWM and DS1103SL_DSP_PWM1) has four PWM channels, and the output is obtained from slave I/O connecter pins. A constant block is connected to the input signal which is used to generate constant duty cycle pulses. The reference signal is defined by the user. The user can set the switching frequency and can develop PWM simulink model. The magnitude of the carrier signal and reference signal limit is 0 to 1. The user can set the switching frequency from 1.25 to 5 MHz and set the duty cycle limit between 0 and 1. In Fig. 6, 0.5 in constant block indicates 50% duty cycle. Pulse output is obtained at slave I/O connector pins 5, 10, 29, and 11. The same procedure is followed for other PWM block (DS1103SL_DSP_PWM3 and DS1103SL_DSP_PWM2). It has three duty cycle connection ports with an additional feature for user to set dead band values for controlling the pulse signals. These PWM pulses are available in slave I/O pins 7–9, and inverted pulses are available in I/O pins 26–28. The pulse output voltage range is 4–5 V.

314

R. Amalrajan et al.

5 Experimental Setup of DC-DC Converter The buck converter is a step-down DC-DC converter which is used for stepping down the input voltage in industry applications. Figure 7 shows the experimental setup of a buck converter and real-time interfacing with dSPACE1103 controller. The buck converter operation depends on the switch position and inductor current. The dSPACE1103 controller is used to generate the switching pulses with a frequency of 20 kHz. Its input DC voltage is 20 V, and a resistance of 25  is connected across the load terminals. The specifications of the converter are given in Table 1. Figure 8 shows the switching frequency signal of buck converter and is observed in mixed digital oscilloscope (MDO_3143). The switching pulse waveform is taken for a duty cycle of 50% using DS1103SL_DSP_PWM3 block with a switching frequency of 20 kHz. The amplitude of the pulse signal is increased to 14.8 V by a TLP 250 driver circuit. Figure 9 shows the closed loop simulation of a buck converter in MATLAB/simulink with PI controller. The simulink file is modified with realtime interface RTI library of dSPACE1103 controller. The buck converter output voltage and duty cycles are measured. In dSPACE, the feedback voltage signal must be with in ±10 V. The output voltage is detected by using a potential divider, and it is feedback through dSPACE1103 chip, via analog-to-digital converter [S1103MUX_ADC_CON1]. For the input voltage ±10 V, the output value of ADC is ±1 V; MUX ADC is used to feed four different analog values. The output analog

Fig. 7 Experimental setup for DC-DC converter

Table 1 List of components

Name of the component

Specifications

IGBT [FGA25N120ANTD]

25 A, 1200 V

Diode [MUR 860]

8 A, 600 V

Inductor

184 μH, 10A

Capacitor

220 μF, 250 V

Resistive load

25 

DSPACE1103 Controller for PWM Control of Power Electronic Converters

315

Fig. 8 PWM waveform for α = 0.5 at 20 kHz switching frequency

Fig. 9 Closed loop simulation of DC-DC converter

voltage of buck converter is converted into digital signal, which is verified in control desk. The digital pulse signal is connected via DAC. For the input range of ±1 V, the output value of the DAC is ±10 V. Figure 10 indicates the experimental results of buck converter viewed in

a

b

Fig. 10 Experimental results for V o = 8 V. a. Set voltage and actual voltage in control desk. b. Duty cycle and output average voltage

316

a

R. Amalrajan et al.

b

Fig. 11 Experimental results for V o = 12 V. a. Set voltage and actual voltage in control desk. b. Duty cycle and output average voltage

dSPACE_1103 control desk and the results captured in Tektronix MDO3024 for the output voltage of 8 V. During the run time, set voltage is varied, and different output average voltages (actual Voltage) are captured in control desk. The results captured in Tektronix MDO3024 are shown in Fig. 11a, b. The reference voltage is set in the control desk directly. The set voltage and acutal voltage calculations are verified in Figs. 10b and 11b using the well-known formula Vo = α ∗ V dc, where ‘α’ is the duty cycle.

6 SPWM Generation The SPWM pulses for power electronic DC-AC converter are generated using simple simulink blocks in RTI library as shown in Fig. 12. For SPWM generation, sinusoidal sgnal is used as a reference signal, and remaining procedures remain same. Figure 13 shows the SPWM pulse generation for DC-AC converter using MATLAB simulink blocks. The sinusoidal reference signal amplitude is 0.8 with a frequency of 50 Hz, and phase delay is 0. Sine wave 1 and sine wave 2 are phase shifted by 120°. The repeating sequence is used to generate carrier switching frequency at 5 kHz. The output is directly connected with DAC which is available in master PCC in RTI library. The MATLAB pulse output is available from the corresponding output pins in DAC. A delay block is used to generate a dead time in simulink for three-phase pulse generation. The switching frequency is 5 kHz, and the dead time is 10 μs for the switches in the same leg which is shown in Fig. 14. The three-phase input sine signal and SPWM pulses are measured by MDO_3143.

7 Experimental Setup for DC-AC Converter Figure 15 shows the overall experimental set up of a three-phase inverter-fed induc-

DSPACE1103 Controller for PWM Control of Power Electronic Converters

Fig. 12 SPWM generation DC to AC converter

Fig. 13. Three-phase SPWM generation using MATLAB simulink

317

318

R. Amalrajan et al.

Fig. 14. Three-phase SPWM waveforms with dead time

Fig. 15 Experimental setup for inverter-fed induction motor drive

tion motor with brake drum arrangements controlled by dSPACE 1103 controller. NI9225 voltage sensor and NI9227 current sensor are used to measure 300 V rms and 5 A rms, respectively. An uncontrolled three-phase rectifier 50 A, 1200 V is used to supply AC-DC input to the three-phase inverter. The DC link capacitor of 450 V, 220 μF is used to minimize the DC ripples. The three-phase SPWM signal generated from the dSPACE1103 chip is connected to the three-phase inverter (IGBT) via DAC

DSPACE1103 Controller for PWM Control of Power Electronic Converters

319

Fig. 16 a Input current. b Output current. c Output voltage

Fig. 17 RPM measurement in simulation

pins. Figure 16 shows the three-phase rectifier input current waveform, output load current, and voltage waveform of a three-phase induction motor load. The waveforms are observed using NI DAC LabVIEW software program in front panel. The speed is sensed by proximity sensor, and the simulink blocks necessary to measure the speed are provided in Fig. 17. The observed proximity sensor output and the corresponding speed waveform are measured in control desk as shown in Fig. 18.

8 Conclusion The overview, working procedure, configuration, and control desk parameter settings of dSPACE1103 controller are presented in simple steps for PWM generation using RTI library. It will be much helpful for young researchers and engineers to learn the working procedure in dSPACE easily. The PWM for duty cycle control of DC-DC converter is presented with experimental results. The closed loop control of DC-DC converter was demonstrated without additional hardware components

320

R. Amalrajan et al.

Fig. 18 Proximity sensor output and speed waveform in control desk

for PI controller and other control circuits. The SPWM for voltage and frequency control of DC-AC converter is demonstrated with induction motor drive using RTI library blocks in dSPACE and MATLAB simulink blocks. The preference of different PWM blocks in RTI library depends upon the user requirement. The dSPACE1103 controller can be used to generate the switching pulses at high frequency (above 100 kHz).

References 1. Chen H, Yen M, Wu Q, Chang K, Wang L (2014) Batteryless transceiver prototype for medical implant in 0.18- m CMOS technology. IEEE Trans Microwave Theory Tech 62(1):137–147 2. Cui S, Soltau ISN, De Doncker RW (2018) A high step-up ratio soft-switching DC-DC converter for interconnection of MVDC and HVDC grids. IEEE Trans Power Electr 33(4): 2986–3001 3. Stieneker SPEM, Soltau N, Rabiee S, Stagge H, De Doncker RW (2015) Comparison of the modular multilevel DC converter and the dual-active Bridge converter for power conversion in HVDC and MVDC grids. IEEE Trans Power Electron 30(1): 124–137 4. Ebrahim B, Okhtay A (2017) A new topology for bidirectional multi-input multi-output buck direct current–direct current converter. Int Trans Electr Energ Syst 27(2):e2254 5. Samson AA, Kumar AS (2014) Design and analysis of fuel cell for standalone renewable energy system. In: IEEE national conference on emerging trends in new & renewable energy sources and energy management (NECT NRES EM), pp 170–175 6. Tani A, Camara MB, Dakyo B, Azzouz Y (2013) DC/DC and DC/AC converters control for hybrid electric vehicles energy and fuel cell. IEEE Trans Ind Inform 9(2):686–696 7. Lee Y, Khaligh A, Emadi A (2009) Advanced integrated bidirectional AC/DC and DC/DC converter for plug-in hybrid electric vehicles. IEEE Trans Veh Technol 58(8):3970–3980 8. Subrahmanya Kumar Bhajana VV, Drabek P, Pramod Kumar A, Design and implementation of a zero voltage transition bidirectional DC-DC converter for DC traction vehicles. Int Trans Electr Energ Syst. https://doi.org/10.1002/2050-7038.2842.

DSPACE1103 Controller for PWM Control of Power Electronic Converters

321

9. Lee H, Yun J (2019) High-efficiency bidirectional buck-boost converter for photovoltaic and energy storage systems in a smart grid. IEEE Trans Power Electr 34(5):4316–4328 10. Han BM, Choi NS, Lee JY (2014) New bidirectional intelligent semiconductor transformer for smart grid application. IEEE Trans Power Electr 29(8):4058–4066 11. Sri Revathi B, Prabhakar M, Longatt FG (2018) High-gain–high-power (HGHP) DC-DC converter for DC microgrid applications: Design and testing. Int Trans Electr Energ Syst 8(2): e2483 12. Chen Q, Liu N, Hu C (2017) Autonomous energy management strategy for solid state transformer to integrate PV-assisted EV charging station participating in ancillary service. IEEE Trans Ind Inform 13(1):258–269 13. Yu X, She X, Zhou X (2014) Power management for DC microgrid enabled by solid-state transformer. IEEE Trans Smart Grid 5(2):954–965 14. de Oliveira Filho HM, de Souza Oliveira D, Praça PP (2016) Steady- state analysis of a ZVS bidirectional isolated three-phase DC-DC converter using dual phase-shift control with variable duty cycle. IEEE Trans Power Electr 31(3):1863–1872 15. Mehrnami S, Mazumder SK, Soni H (2016) Modulation scheme for three-phase differential´ inverter. IEEE Trans Power Electr 31(3):2654–2668 mode Cuk 16. Liang Z, Lin X, Kang Y, Gao B, Lei H (2018) Short circuit current characteristics analysis and improved current limiting strategy for three-phase three-leg inverter under asymmetric short circuit fault. IEEE Trans Power Electr 33(8):7214–7228 17. Tamyurek B (2013) A high-performance SPWM controller for three-phase UPS systems operating under highly nonlinear loads. IEEE Trans Power Electr 28(8):3689–3701 18. Song Y, Wang B (2013) A survey on reliability of power electronic systems. IEEE Trans Power Electr 28(1):7214–7228 19. Chinmaya KA, Singh GK, Experimental analysis of various space vector pulse width modulation (SVPWM) techniques for dual three-phase induction motor drive. Int Trans Electr Energ Syst. https://doi.org/10.1002/etep.2678 20. Sun P, Liu C, Lai J (2012) Three-phase dual-buck inverter with Unified Pulse width Modulation. IEEE Trans Power Electr 27(3):7214–7228 21. Housheng Z, Yanlei Z (2010) Research on a novel digital photovoltaic array simulator. In: International conference on intelligent computing technology and automation. pp 1077–1080 22. Lu DDC, Nguyen QN (2012) A photovoltaic panel emulator using a buck-boost DC/DC converter and a low cost micro-controller. Sol Energy 86(5):1477–1484 23. Fajri P, Lee S, Anand Kishore PV, Ferdowsi M (2016) Modeling and integration of electric vehicle regenerative and friction braking for motor/dynamometer test bench emulation. IEEE Trans Veh Technol 65(6): 4264–4273 24. Subrata B, Arnab G, Sanjeevikumar P (2019) Modeling and analysis of complex dynamics for dSPACE controlled closed-loop DC-DC boost converter. Int Trans Electr Energ Syst. https:// doi.org/10.1002/etep.2813(availableonlinefromJan 25. Gebregergis A, Pillay P (2010) Implementation of fuel cell emulation on DSP and dSPACE controllers in the design of power electronic converters. IEEE Trans Ind Appl 46(1):285–294 26. Sarker MR, Mohamed R (2019) dSPACE controller-based enhanced piezoelectric energy harvesting system using PI-lightning search algorithm. IEEE Access 7:3610–3626 27. Azharuddin SM, Vysakh M, Thakur HV, Nishant B, Sudhakar Babu T, Muralidhar K, Paul D, Jacob B, Balasubramanian K, Rajasekar N (2014) A near accurate solar PV emulator using dSPACE controller for real-time control. Energy Proc 61: 2640–2648 28. Meshram RV, Bhagwat M, Khade S, Wagh SR, Stankovi AM, Singh NM (2019) Port-controlled phasor hamiltonian modeling and IDA-PBC control of solid-state transformer. IEEE Trans Control Syst Technol 27(1):349–356

322

R. Amalrajan et al.

29. Hac T, Karaman S, Kural E, Öztürk ES, Demirci M, Güvenç BA (2006) Adaptive headlight system design using hardware-in-the-loop simulation. In: IEEE International conference on control applications, pp 915–920 30. yThomas A, Anakwa AW (2016) dSPACE DS1103 control workstation tutorial and DC motor speed control tutorial 2(5):99–110

Performance and Reliability Analysis of 13-Level Asymmetrical Inverter with Reduced Devices Abeera Dutt Roy

and Chandrahasan Umayal

Abstract This paper presents an asymmetrical multilevel inverter topology, which employs reduced power electronic devices to produce 13 levels at the output voltage. This reduction in the components results in the curtailment of the size, cost, switching and conduction losses. A comparative study of the proposed topology with the recent ones is presented. Nearest level control (NLC) is utilized for the generation of gating pulses. A reliability analysis is performed using Markov reliability calculation, and the efficiency of the overall system is improved. Simulation and experimental results are provided for resistive and inductive loads. Keywords Multilevel inverter · Asymmetrical · Nearest level control · Reliability

1 Introduction The advantages of MLIs lie in its ability to produce lesser blocking voltages, reduced electromagnetic interference (EMI) and lower THD of the semiconductor devices [1, 2]. The classical topologies of multilevel inverters are classified as neutral point clamped (NPC), cascaded H-bridge (CHB) and flying capacitors (FCs) [3–5]. Among all these topologies, cascaded multilevel inverter is adapted for high-power applications due to its modular structure which allows higher voltage execution by using low-voltage semiconductor devices. The advantage of multilevel inverters [MLIs] lies in its ability to produce greater voltage levels by utilizing reduced component count [6]. Therefore, many topologies are reported in the literature with reduced component count [7–12]. Various asymmetrical topologies utilize full bridge inverter for the achievement of negative voltage levels at the output [13]. The drawback of this topology is the high stress that is produced in the power devices of the H-bridge. So, the literature has several architectures which have the ability to produce the negative voltage levels at the output side without using H-bridge inverter [14, 15]. A. D. Roy (B) · C. Umayal School of Electrical Engineering, Vellore Institute of Technology, Vellore, Chennai 632014, Tamil Nadu, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_25

323

324

A. D. Roy and C. Umayal

This paper deals with a 13-level inverter topology with fourteen switches and two unequal DC sources. Since the DC source values are unequal, this topology is an asymmetrical variant. A comparative study is done between the proposed and the recent structures. NLC technique is used for the generation of the gating pulses. A prototype is developed in the laboratory and tested for various loads for the validation of the experimental results.

2 Proposed Topology Figure 1 shows the circuit diagram of the proposed 13-level inverter topology. It consists of eight unidirectional and three bidirectional switches which contribute in the creation of six positive and six negative levels and a zero level. It also utilizes two unequal DC sources to produce the desired waveform. The switching states of this structure are shown in Table 1. It is noticed that switches S 1 , S 7 and S 5 and S 9 cannot be turned on at the same time in order to prevent short circuit between the voltage sources. For the creation of +1V dc , the switches S 1 , S 8 and S 9 are turned on, and for the creation of 2V dc , the switches S 5 , S 7 and S 8 are turned on. The maximum blocking voltage for each switch is determined, and then, the total of all the blocking voltages across all the switches is given as TSV which is given by in Eq. (3).

Fig. 1 Circuit of the proposed topology

Performance and Reliability Analysis of 13-Level … Table 1 Switching states of the 13-level topology

325

Levels

‘ON’ state switches

1

S1 , S8 , S9

2

S 5 , S 7 ,S 8

3

S1 , S5 , S8

4

S 1 , S 4 , S 9 , S 10 , S 11

5

S 4 , S 5 , S 7 , S 10 , S 11

6

S 1 , S 4 , S 5 , S 10 , S 11

−1

S3 , S5 , S7

−2

S1 , S3 , S9

−3

S3 , S7 , S9

−4

S 3 , S 6 , S 7 , S 11

−5

S 2 , S 3 , S 9 , S 10

−6

S 2 , S 3 , S 6 , S 10 , S 11

The number of switches is given by Eq. (1)  Nl − 2 +1 = 14 12 

N sw

(1)

The number of DC sources is given by Eq. (2)  Ndc = 2

 Nl − 2 +1 12

(2)

The total standing voltages (TSV) are given by Eq. (3)  Nl − 2 +1 TSV = 32 12 

(3)

3 Modulation Technique The NLC methodology is utilized for producing the switching pulses at the fundamental switching frequency. Since it is a low switching frequency method, it generates reduced switching losses and quite simple to implement for prototyping [14]. It produces higher harmonics for lower levels of the output waveform. The switching angle calculation is given by the following Eq. (4): αi = sin−1

i − 0.5 for i = 1, 2, 3, . . . , n n

(4)

326

A. D. Roy and C. Umayal

The THD is determined with the calculated values of the switching angles and is given by Eq. (5):  THD =

π 2 n2 8



π

  n

n−1

i=0 (2i+1)αi+1 −

n

(

i=1

cos αi )

2



4

(5)

i=1 cos(αi )

4 Comparative Study Table 2 shows the comparison of the proposed topology with the recent ones in terms of the number of switches and DC sources utilized. Figure 2 shows the number of Table 2 Comparison of the proposed and recent topologies

Topology [14] [15] [16] [17] [18] [19] Proposed

Switches   l −2 10 N12 +1   l −3 12 N16 +1   6 Nl4−2 + 1   6 Nl6−2 + 1   l −3 22 N16 +1   8 Nl8−2 + 1   Nl −2 8 +1

Fig. 2 Number of switches versus number of levels

DC sources   l −2 4 N12 +1   l −3 4 N16 +1   Nl −2 4 +1   Nl −2 6 +1   l −3 2 N16 +1   Nl −2 8 +1   2 Nl8−2 + 1

Negative level Inherent Inherent Inherent Inherent With H-bridge With H-bridge Inherent

Performance and Reliability Analysis of 13-Level …

327

Fig. 3 Number of switches versus number of levels

switches versus the number of levels used. The number of switches used is the highest in topology [18], and the least number of switches is used by topology [19]. Figure 3 displays the utilization of the DC sources where it is observed that the topology in [19] has the least number of sources used, while the topology in [14] has the highest usage. Table 2 also highlights the fact this topology is able to produce the negative levels at the output without using H-bridge like other topologies [14–17].

5 Reliability Analysis One of the important features of any MLI or power electronic converter is its longer lifespan, increased efficiency and improved quality of the output waveform. And a converter with higher lifespan is expected to possess higher reliability which in turn is defined as possibility of any device fulfilling its expected operation for a fixed time duration within its safe operating area. It is important to determine the reliability of every component of the system. Failure rate: It plays an important role in deciding the reliability of the system. It is defined as the ratio of the rate of occurrence of the failures of a component for a fixed duration to the total time of the operation. It is written as FIT, where one FIT means the occurrence of one failure in 109 operation hours. According to MIL-HDBK-217F standard (Defense, 1991), the probability distribution of failure of electronic devices is given by Eq. (6), f (t) = λ eλt (t > 0)

(6)

where λ is the failure rate. Then the reliability function is denoted by R(t) and shown in Eq. (7)

328

A. D. Roy and C. Umayal t

R(t) = 1 − ∫ f (t)dt = e−λt

(7)

0

MTTF: When the system starts, the average time taken by the first failure of the component is evaluated in hours and it is an important parameter of a device. It is expressed in Eq. (8) ∞ MTTF =

R(t)dt = e−λt =

0

1 λ

(8)

One of the method for reliability analysis is part stress method where the parameters of the converters are measured. It requires information about every part stress the levels for different operating conditions. The total failure rate of the system using this method is calculated by using Eq. (9) λsystem= Np

i=1

(9)

λpart

where N p is the number of the part. The failure rate of switch, diode and capacitor is calculated by using the Eqs. (10), (11), and (12): λPS = λBS ∗ πQ ∗ πA ∗ πE ∗ πT

(10)

λPD = λBD ∗ πQ ∗ πE ∗ πC ∗ πT ∗ πS

(11)

λCP = λCB ∗ πCV ∗ πQ ∗ πE

(12)

where the superscripts ‘S’, ‘D’ and ‘C’ refers to switch, diode and capacitor respectively.The base failure rate λBS and λBD is constant and equal to 0.012 and 0.064 respectively. The capacitor failure rate is calculated as  λCB

= 0.00254 ×

S 0.5





TA + 273 5 + 1 × exp 5.09 × 378

(13)

where S is equal to the ratio of operating voltage to rated voltage. The temperature factor (πT ) is calculated as follows πTS

= exp −1925 ×

πTD = exp −1925 ×

1 1 − T j + 273 298 1 1 − T j + 273 293



(14)

(15)

Performance and Reliability Analysis of 13-Level …

329

where Tj is the device junction temperature and the stress factor (πS ) and capacitance factor (πCV ) are calculated as follows: πs = Vs243

(16)

where Vs is equal to the ratio of operating voltage to rated voltage and πCV = 0.34 × C 0.12

(17)

where C is the capacitance in microfarad. The values of quality factor (πQ ), environment factor (πE ), application factor (πA ) and construction factor (πC ) respectively are equal to 5.5, 6,10 and 1. The Markov reliability calculation is frequently used for the determination of systems which are either discrete or continuous. Since power electronic converter is discrete, so this method can be applied to measure its reliability. The steps to be followed are shown below, and its flowchart is shown in Fig. 4a. 1. 2. 3. 4.

The power losses of all the switches are computed. The junction temperatures of each  device are obtained. C for all the present components. The factors πT , πS , πCV andλ 

B are calculated The aggregate failure rates λPS , λPD , λCP are determined for every component and then the overall reliability is calculated and the same is shown in Fig. 4a. The parameters used for the simulation are shown in Table 3. By using these

Fig. 4 a Flow chart of reliability analysis. b Reliability curve of proposed topology

Table 3 Parameters employed for simulation V dc = 150 V

R = 80 

f sw = 50 Hz

330

A. D. Roy and C. Umayal

Table 4 Power loss (W) of components S1 = 31.4

S2 = 36.64

S3 = 37.27

S9 = 0.154

S 10 = 0.154

S 11 = 0.443

S 4 = 31.4

S5 = 0.413

S6 = 0.456

S7 = 0.399

S8 = 0.443

Table 5 Failure rate of components S1 = 6.5512

S2 = 6.5512

S 10 = 6.538

S 11 = 6.538

S3 = 6.556

S4 = 6.556

S5 = 6.553

S6 = 6.556

S7 = 6.552

S8 = 6.555

S9 = 6.538

parameters, the losses obtained for each switch are shown in Table 4. Thereafter, the failure rate of each component is displayed in Table 5. Thus the reliability function and MTTF are shown in Eq. (13) where λ = λS,11 = 72.06 and MTTF are given by Eq. (14) and the reliability curve of this topology is shown in Fig. 4b. R(t) = e−λt = e−72.062t

(18)



1 × 106 h = 0.0138 × 106 h MTTF = λ

(19)

6 Simulation Results The proposed topology is simulated on an R2015b MATLAB platform. A output voltage of 13 levels is produced for load of R = 50  and L = 32 mH. The NLM technique is used for the production of gating pulses. Figure 5 displays output voltage and current for R load. It is seen that both the waveforms have zero phase difference between them. Figure 7 depicts the harmonic spectrum of output voltage and THD is 4.78%. For RL load, the output voltage and current waveforms are shown in Fig. 6. A nearly sinusoidal current waveform is observed since the RL load functions like low pass filter. The % THD of output voltage is 5.30% and shown in Fig. 8, whereas Fig. 9 displays the % THD of output current as 0.89%.

7 Experimental Results The NLM technique is established by using the controller ATMEGA32P, and the switching devices used are 25N120. The output voltage and current waveforms are

Performance and Reliability Analysis of 13-Level …

Fig. 5 Output voltage and current for R load

Fig. 6 Output voltage and current for RL load Fig. 7 Output voltage spectrum for R load

331

332

A. D. Roy and C. Umayal

Fig. 8 Output voltage spectrum for RL load

Fig. 9 Output current spectrum for RL load

shown in Figs. 10 and 11 for R and RL loads, respectively. The RMS value of output voltage and current is 109.46 V and 4.5 A, respectively. And the output voltage and current for RL load were observed as 106.5, 4.7 A, respectively. The voltage THD of 5.8 and 6.4% for R and RL load is shown in Figs. 12 and 13, respectively. And the current THD of 1.2% is depicted in Fig. 14. A phase difference is observed between output voltage and current due to the inductive nature of the load. The power quality analyzer Fluke 435B is used for observing the waveforms and the harmonic content. Fig. 10 Output voltage and current for R load

Performance and Reliability Analysis of 13-Level … Fig. 11 Output voltage and current for RL load

Fig. 12 Output voltage spectrum for R load

Fig. 13 Output voltage spectrum for RL load

333

334

A. D. Roy and C. Umayal

Fig. 14 Output current spectrum for RL load

8 Conclusion In this paper, an asymmetrical multilevel inverter is presented which produces 13 levels at the output using 14 power electronic switches and two DC sources. The comparative study provided between the proposed and the recent topologies showed the reduction in the power devices. The nearest level modulation technique is used for the production of gating pulses. Reliability analysis is performed using the Markov chain methodology, and the reliability curve for the same is plotted. The performance of this topology is validated for various loads.

References 1. Rodriguez J, Lai J-S, Peng FZ (2002) Multilevel inverters: a survey of topologies, controls, and applications. IEEE Trans Ind Electron 49(4):724–738 2. Franquelo L, Rodriguez J, Leon J, Kouro S, Portillo R, Prats M (2008) The age of multilevel converters arrives. IEEE Ind Electron Mag 2(2):28–39 3. Malinowski M, Gopakumar K, Rodriguez J, Pérez MA (2010) A survey on cascaded multilevel inverters. IEEE Trans Ind Electron 57(7):2197–2206 4. Peng FZ (2001) A generalized multilevel inverter topology with self-voltage balancing. IEEE Trans Ind Appl 37(2):611–618 5. Nabae A, Takahashi I, Akagi H (1981) A new neutral-point-clamped PWM inverter. IEEE Trans Ind Appl IA-17(5):518–523 6. Colak I, Kabalci E, Bayindir R (2011) Review of multilevel voltage source inverter topologies and control schemes. Energy Convers Manag 52(2):1114–1128 7. Hasan MM, Abu-Siada A, Dahidah MSA (2018) A three-phase symmetrical DC-Link multilevel inverter with reduced number of DC Sources. IEEE Trans Power Electron 33(10):8331– 8340 8. Oskuee MRJ, Karimi M, Ravadanegh SN, Gharehpetian GB (2015) An innovative scheme of symmetric multilevel voltage source inverter with lower number of circuit devices. IEEE Trans Ind Electron 62(11):6965–6973

Performance and Reliability Analysis of 13-Level …

335

9. Dhanamjayulu C, Meikandasivam S (2018) Implementation and comparison of symmetric and asymmetric multilevel inverters for dynamic loads. IEEE Access 6:738–746 10. Samadaei E, Kaviani M, Bertilsson K (2018) A 13-levels Module (K-Type) with two DC sources for multilevel inverters. IEEE Trans Ind Electron 1-1 11. Ray S, Gupta N, Gupta RA (2018) Prototype development and experimental investigation on cascaded five-level inverter based active filter for large-scale grid-tied photovoltaic. Int J Renew Energy Res (IJRER) 12. Karthikeyan D, Vijayakumar K, Jagabar SMA (2019) Generalized cascaded symmetric and level doubling multilevel converter topology with reduced THD for photovoltaic applications. Electronics 8(2):161 13. Ebrahimi J, Babaei E, Gharehpetian GB (2012) A new multilevel converter topology with reduced number of power electronic components. IEEE Trans Ind Electron 59(2):655–667 14. Samadaei E, Gholamian S, Sheikholeslami A et al (2016) An envelope type (E-type) module: asymmetric multilevel inverters with reduced components. IEEE Trans Ind Electron 63(11):7148–7156 15. Samadaei E, Sheikholeslami A, Gholamian SA, Adabi J (2018) A square T-type (ST-type) module for asymmetrical multilevel inverters. IEEE Trans Power Electron 33:987–996 16. Vahedi, Al-Haddad K (2016) PUC5 inverter—a promising topology for single-phase and threephase applications. IECON 2016—42nd annual conference of the IEEE industrial electronics society, Florence, pp 6522–6527 17. Vahedi H, Al-Haddad K (2016) Real-time implementation of a seven-level packed Ucell inverter with a low-switching-frequency voltage regulator. IEEE Trans Power Electron 31(8):5967–5973 18. Zamiri E, Vosoughi N, Hosseini SH, Barzegarkhoo R, Sabahi M (2016) A new cascaded switched-capacitor multilevel inverter based on improved series-parallel conversion with less number of components. IEEE Trans Industr Electron 63(6):3582–3594 19. Ye Y, Cheng KWE, Liu J, Ding K (2014) A step-up switched-capacitor multilevel inverter with self-voltage balancing. IEEE Trans Industr Electron 61(12):6672–6680

Lighting Design for a Campus Hostel Building Using Single-Stage Efficient Buck LED Driver R. Srimathi , Kishore Eswaran, S. Hemamalini , and V. Kamatchi Kannan

Abstract Lighting is an important feature for the interior of a room as it produces a good state of mind, atmosphere, and beautiful appeal in living space. Light and lighting fixture create a pleasant and safe habitat besides adding style to the interior of a room. Moreover, as lighting consumes nearly 35% of the total electricity consumed in a building, LED drivers are proposed to increase energy efficiency. Lighting design is mandatory for the placement of lights to have good illumination, thereby lighting design is done for a two-bedded hostel room with fluorescent and LED lights and their performance is analyzed. A phase-shifted buck converter (PSBC) LED driver is designed and experimented for 24 and 6 W to retrofit the existing fluorescent lights in the two-bedded room. The results show that the output lumens and efficiency of PSBC LED driver are higher than conventional LED driver (CBC). Based on the PSBC experimental results, a lighting design is done using DIALux software for the existing room with fluorescent lights and a retrofitted room using PSBC and LED lights. The simulation results show that PSBC driver provides good illumination and has better energy savings. The estimated installation cost, energy savings, and energy efficiency for the PSBC driver and a conventional buck driver also prove that PSBC driver is a better alternative for energy savings. Keywords Phase-shifted buck converter (PSBC) · Digital controller · Lighting design · DIALux software · Energy savings and energy efficiency

1 Introduction Lighting plays a major role in the interior of a building. Accent, task, and general lighting are the types of lighting. General or ambient lighting [1] is for indoor and parking lots, whereas task lighting [2] is preferred for important tasks such as surgical, R. Srimathi (B) · K. Eswaran · S. Hemamalini Vellore Institute of Technology, Vellore, Chennai 600127, Tamil Nadu, India e-mail: [email protected] V. Kamatchi Kannan EEE Department, BIT, Sathyamangalam, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_26

337

338

R. Srimathi et al.

reading, and in factories. Accent lighting [3] is used for decorative purpose and in landscaping. Though there are different lighting types, lighting in a building can be classified as interior and exterior lighting. The use of artificial lighting sources such as tungsten lamps, fluorescent, sodium-vapor lamp, and LEDs is common to bring aesthetic effects in a building. However, LEDs are well known for its luminous efficiency and long lifetime than other lighting sources [4]. Buildings in general are classified as residential, commercial, and agricultural buildings. Campus buildings are a type of commercial buildings. Most of the loads connected in campus building are DC-based loads. This includes computers, telephones, televisions, LED lights, etc. These devices are plugged directly into an alternating current (AC) outlet which is then converted from AC to DC [5]. The losses associated with the conversion of AC to DC for these loads result in significant power loss. Therefore to reduce the power loss, these loads can be supplied by low voltage DC (LVDC) supply [6]. Lighting consumes nearly 35–45% of energy in campus buildings. Based on the nature of power supply, LED drivers are classified as AC and DC LED drivers [7]. The power stage of an AC lighting system is shown in Fig. 1. The LED backlight system consists of three cascaded power conversion stages such as power factor control circuit (PFC), isolated DC-DC converter, and LED driver circuit. The PFC circuit is a rectifier circuit which is also used to regulate the power factor of the source closer to unity, and the DC-DC converter converts variable DC to fixed DC of 24 V for the LED lights. The output of the isolated DCDC converter is 24 V DC. The LED driver circuit produces the required DC voltage from the isolated DC-DC converter to drive the LED strings [8]. The power stage of an AC lighting consists of multiple stages of power conversion such as AC to DC and DC-DC for the lighting source as in Fig. 1. The cost of the AC driver circuit is higher than a DC driver. This is due to more number of power conversion stages in AC. The inefficiency of each conversion is also multiplied. Hence, to conserve energy, single-stage DC drivers are preferred for buildings [9]. But one of the issues regarding the adoption of DC is that while devices may be internally powered by DC they are generally manufactured and sold with AC plugs/power supplies. In a new construction, this issue could be addressed by simply procuring new devices that can accommodate direct DC power. With existing buildings, it is necessary to retrofit the existing AC devices by DC thereby to accommodate energy conservation in buildings [17]. The aim of this paper is to design and develop a single-stage phaseshifted buck converter (PSBC) [10] as LED driver for campus hostel building that could be used for interior lighting. The PSBC driver has four modes of operation and is designed for 24 W output power with an input voltage of 24 V. The converter is

Fig. 1 Power stage of AC lighting system

Lighting Design for a Campus Hostel Building …

339

also experimented in closed loop for load variations. Moreover, the lighting design for a two-bedded hostel room is done using DIALux software with PSBC driver. The installation cost, energy estimation, and energy savings for the campus building is also calculated and projected in this paper.

2 Phase-Shifted Buck Converter (PSBC) The PSBC DC driver is a buck converter with two parallel switches (S 1 , S 2 ), a diode (D), an inductor (L), and a capacitor (C). The circuit diagram for PSBC is shown in Fig. 2. The driver is operated in four different modes, and the equivalent circuit of the operating modes is shown in Fig. 3. In Mode 1, switch (S 1 ) is turned ON and the current flows from the source, through S 1 , L and C to the load. During this period, the inductor and capacitor are charged and the diode is turned OFF. In Mode 2, the switches S 1 ; S 2 are turned OFF and the diode starts to conduct. The current flows via D, L, and C to the load as the stored energy in the inductor and capacitor is discharged to the load. The operation of Mode

Fig. 2 Phase-shifted buck converter. a Mode 1—S 1 —ON and S 2 OFF. b Mode 2—S 1 and S 2 are OFF. c Mode 3—S 2 —ON and S 1 —OFF. d Mode 4—S 1 and S 2 are OFF

(a) Mode 1 – S1 – ON and S2 OFF

(b) Mode 2 – S1 and S2 are OFF

(c) Mode 3 – S2- ON and S1- OFF Fig. 3 Equivalent circuit for modes of operation of PSBC

(d) Mode 4 - S1 and S2 are OFF

340

R. Srimathi et al.

Fig. 4 Ideal waveforms of PSBC

3 is similar to Mode 2 except that switch S 1 is OFF instead of S 2 . In the same way, Mode 4 and Mode 2 have similar operations. The ideal waveforms for the operating modes are shown in Fig. 4.

3 Design of PSBC Driver The voltage gain and current gain equation of PSBC driver are given in (1) and (2).

Lighting Design for a Campus Hostel Building …

341

Fig. 5 Schematic for closed-loop control of PSBC

Vo =m∗D Vin

(1)

Iin =m∗D Io

(2)

where I in and I o are the input and output currents of the converter, respectively. In (1) and (2), the duty ratio of the switches are assumed to be equal, i.e., Dl = D2 = D. The design equation for L and C are derived from the modes of operation and are given in (3) and (4). L=

m D ∗ Vin ∗ (1 − m ∗ D) IL ∗ 2 ∗ f s

(3)

C=

m D ∗ Vin ∗ (1 − m ∗ D) 32 ∗ V0 ∗ L ∗ f s2

(4)

where m is the number of parallel switches, D is the duty ratio, and f s is the switching frequency. The converter is operated in closed loop by designing a digital current controller [11, 12, 16] using PIC microcontroller. The closed-loop control circuit diagram for PSBC converter is shown in Fig. 5. The design specifications of PSBC driver for 24 and 6 W are given in Table 1.

4 Lighting Design for a Two-Bedded Hostel Room VIT Campus hostel building is a multi-story building with 12 floors. There are 252 non-AC two-bedded rooms in C block boys hostel. There is a main light of 48 W fluorescent tube light and a 12 W fluorescent mirror lamp in the two-bedded hostel room. When a fluorescent lamp is replaced by LED light by retrofit, the power consumption

342 Table 1 Design specifications of PSBC

R. Srimathi et al. Parameters

24 W PSBC

Input voltage (V in )

24 V

6W PSBC

Output voltage (V o )

12 V

Duty ratio (D)

0.5

Operating frequency (f s )

25 kHz

Input current (I in )

1A

0.25 A

Output current (I o )

2A

0.5 A

is reduced to half [13, 15]. Therefore, the PSBC driver is designed for 24 and 6 W to retrofit the fluorescent lights. The two-bedded room is equipped with two wooden study tables, two beds, and cupboards. The existing two-bedded hostel room has a measurement of 4.5 m * 4.5 m * 4 m with a light colored wall. The total area of this room is 20.25 m2 . The two-bedded hostel room is taken into consideration for lighting design and is simulated using software tool DIALux evo 8.1 [14]. Lighting design is the application of lighting which relies on a combination of specific scientific principles, established standards and conventions, and a number of human and natural factors. In present years, the field of lighting has been enhancing with two prominent forces, energy efficiency, and lighting quality. Therefore, selection of the room lighting helps us to achieve a perfect balance between style, comfort, and cost. Transmittance and reflection of the light in a room depend on the properties of the material involved. Room has a reflection factor of 40–60% [18], a glass window with transmission factor of 92%, and four walls with reflection property of 50%. Factors considered for lighting design are luminous efficacy, unified glare rating (UGR), installed load efficacy ratio (ILER), spacing of luminaries, and color rendering index (CRI).

4.1 Design Calculations for Illuminance and Load Efficacy Ratio The factors of lighting design for 24 W PSBC for the two-bedded room are computed and the calculations for estimating ILER are given below. Room Index The room index is calculated using (5) Room Index =

l ∗b h ∗ (l + b)

(5)

The length (l), breadth (b), and height (h), and area of the room are 4.5, 4.5, 3.75, and 20.25 m2 , respectively. The two-bedded room has 0.642 as room index. If the room

Lighting Design for a Campus Hostel Building …

343

index is less than one, there should be 8–9 measuring points in the room or else there should be more number of measuring points to calculate the lumens distribution in the room. Average illuminance and Installed Efficacy Ratio (ILER) The average illuminance (E av ) is given by (6). E av =

E1 + E2 + E3 + · · · En 8

(6)

The correction factor of LED and fluorescent lamp is 1 and 1.08 [18], respectively. The net average illuminance (NE av ) is calculated by the product of E av and correction factor. As a result, NE av for LED and fluorescent are 6598.625 and 3699.14 lm, respectively. Available lumens on the measuring plane (ϕ) are calculated using (7). ϕ = N E av ∗ area of the room

(7)

For LED and fluorescent lamps, ϕ is 133622.16 and 490.48, respectively. The installed load efficacy (ILE) is estimated using (8). ILE =

Total lumens on measurement plane Wattage for Lighting Circuit

(8)

From (8), the ILE for LED and fluorescent lamps are 2889.95 and 361.95 lm/W, respectively. The target ILE is given from BEE code is 36–40 lumens per meter for the proposed room. This is as per the application and CRI of the LED lamp. ILER is estimated using (9). ILER =

Installed Load Efficacy Target Installed Load Efficacy

(9)

The ILER for LED and fluorescent tube (48 W) is 3.56 and 1.68, respectively. As per BEE code [18], if ILER is above 0.75, the performance of the light is good and energy efficient. It is evident from the computational results that both the main and the mirror light are energy efficient. The results for 6 W LED is also computed in a similar manner using (6)–(9) and is given in Table 2. The structure of the room is shown in Fig. 6.

5 Experimental Results The PSBC driver is designed for 24 and 6 W for the two-bedded hostel room. The converter is implemented in hardware for specifications as mentioned in Table 1. The part number of the components is mentioned in Table 3 along with its cost. The

344 Table 2 Computational results

R. Srimathi et al. Parameters

Results

Room type

Main light Mirror light

Activity type

Reading, writing

Number of lamps

2

Length (m)

4.5

Breadth/width (m)

4.5

No. of illuminance points taken (lux) 8

5

Average room illuminance (W)

3425.13

493.5

Estimated circuit power (lm/W)

24

6

Installed lighting efficacy (lm/W)

2889.95

82.2

Target lighting efficacy (p.u)

810

810

Installed lighting ratio

3.5

0.1

Fig. 6 DIALux model of two-bedded hostel room

hardware prototype of the converter and the experimental set up for PSBC driver for 24 W are shown in Fig. 7. The gate pulses, voltage waveforms across the diode and inductor, the output waveforms, and the regulated waveforms for 24 W are shown in Fig. 8a–d. It is observed from Fig. 8d, b, that the gate pulses for switches S 1 and S 2 are phase shifted with each other, whereas the voltage across inductor (V L ) and diode (V d ) is complementary in nature. The output current is regulated for load variations as seen from Fig. 8d. The experimental results of 6 W PSBC are shown in Fig. 9. The power switches are operated in phase shift and the sequence of gating signal in Fig. 9a whereas the output waveforms for 6 W are shown in Fig. 9b. The input power and output power are measured from the experimental results to calculate the efficiency. The efficiency of the PSBC driver for variation in duty ratio is shown in Fig. 10. From the experimental results, it is evident that the efficiency of PSBC is higher by 5% than CBC. Using lux meter (LX-101A), the lumen output for 6 and 24 W is measured as

Lighting Design for a Campus Hostel Building …

345

Table 3 Part numbers used and cost of the prototype Components

CBC

PSBC

No.

Cost [12 W] (INR)

Cost [6 W] (INR)

No.

Cost [12 W] (INR)

Cost [6 W] (INR)

MOSFET (IRL1520NPbF)

2

120

120

4

240

240

Diode (SB5H0)

2

75

75

2

75

75

Inductor DMT2-796-1.5 DMT3-1439-1.5 RFC0807B-275 DMT2-796-1.5

2

350

600

2

265

400

Capacitor

2

10

10

2

10

10

Microcontroller (pic18f4550)

2

200

200

2

200

200

LED and sink



60

60



60

60

PCB board

2

200

200

2

200

200

Over heads (INR)

450

450

Total cost (INR)

2655

2675

(a) Hardware Board for PSBC.

(b) Experimental Setup.

Fig. 7 Prototype board and experimental setup of 24 W PSBC. a Hardware board for PSBC. b Experimental setup

540 and 4080 lm from the experimental output. Though the cost and components are higher in PSBC than CBC, the efficiency of the PSBC is higher than CBC. Therefore, PSBC will be a better alternative to retrofit the existing fluorescent lights in campus buildings.

346

R. Srimathi et al.

(a) Voltage and current of diode.

(c) Output Waveforms.

(b) Voltage across diode and inductor.

(d) Dynamic Response for change induty ratio.

Fig. 8 Experimental waveforms of 24 W PSBC. a Voltage and current of diode. b Voltage across diode and inductor. c Output waveforms. d Dynamic response for change in duty ratio

(a) Gate pulses and Switch currents.

(b) Output Waveforms.

Fig. 9 Output waveforms of 6 W PSBC. a Gate pulses and switch currents. b Output waveforms

Lighting Design for a Campus Hostel Building …

347

Fig. 10 Efficiency plot for 24 W PSBC

6 Simulation of PSBC Driver in DIALux Software The proposed buck LED driver is simulated in DIALux evo 8.1 environment and it is compared with fluorescent lamps for the hostel room. The room is equipped with a CFL lamp of Philips make (TMS022 1xTL-D58W HFS + GMS022 R) and a mirror lamp (TCS460 1xTL5-13 W HFP D8) with clearance height of 2.800 m. The reflection factors of the room are considered as 70% for the walls and 15% for the oor with a light loss factor of 0.80. The room is retrofitted with one main LED lamp (4MX850 G3 491 1xLED40S/840 PSD DA) and one LED mirror lamp (LL523XPSD ELD1 EM 1xLED100S/830 DA25N) of 4000 and 600 lm in DIALux environment. The main lamp is mounted on the ceiling height of 1.6 m from the floor, and mirror lamp is mounted just above the mirror. Lighting power density of oor area of room is 3.50 W/m2 . Simulation results are mainly observed by polar intensity diagram. Polar curves refer to the spread of light from the source of illumination from a 3-D space on to a 2-D medium. This curve has two components namely vertical plane through horizontal angle (90°–270° plane—C90/270) and horizontal cone through vertical angle (0°–180° plane C0/180). The spread of light in candela in any particular direction should be good in both the planes. Moreover, the luminous efficacy (LOR) of the lamp should also be high. The polar intensity results of PSBC LED-based and CFL-based luminaries are shown in Fig. 11. From Fig. 11, it is seen that there are two lines radiated from the center which is of two planes namely C90/270 (side view) and C0/180 (front view). Light distribution is measured from the center point of the luminaries. From Fig. 11a, b, the proposed LED driver main light (24 W) outspreads 600–650 cd for 45° spread with an LOR of one, whereas for fluorescent main lamp for a spread of 45° the light distribution is only 200–300 cd with an LOR of 0.81. Similarly, the LED mirror light of (6 W) and the fluorescent mirror light are shown in Fig. 11c, d, and the flux outspreads for 300–660 and 100–400 cd, respectively. Again, the LOR of mirror LED is one and fluorescent LED is 0.74, thereby it is evident that the LED lamp has higher output flux than fluorescent lamp and more

348

R. Srimathi et al.

(a) LED – main light

(b) Fluorescent - main light

(c) LED – mirror light

(d) LED - Fluorescent light

Fig. 11 Polar intensity diagram of LED and CFL. a LED—main light. b Fluorescent—main light. c LED—mirror light. d LED-fluorescent light

(a) Simulation diagram of LED

(b) Simulation diagram of CFL

Fig. 12 3-D Simulation diagram of LED and CFL luminaries. a Simulation diagram of LED. b Simulation diagram of CFL

efficient too. It is also observed from Fig. 11, that ILER and illumination efficiency of the proposed PSBC LED luminarie are higher than fluorescent lamp luminarie as it covers maximum angle (frontal and side angle) with higher illuminance. The total lamp luminous flux and obtained luminaire luminous flux of both main and mirror light for CFL and LED lamps simulated in DIALux are shown in Fig. 12. It is observed that, the light output ratio of CFL is 77% only when the LED has 100% output ratio with higher luminous efficacy.

7 Energy Estimation and Savings The energy estimation for the campus hostel building is focussed in this section. Consider, if the rooms main light is switched ON for approximately 8 h a day and the mirror light is switched ON for one hour in a day, then total energy consumed in one year is 3200 h. The energy saving and payback period for main light in a room are given in Table 4. From Table 4, it is evident that, there is considerable energy savings with LED. Moreover, net pro_t with energy savings is INR.1.73 lakhs per annum for

Lighting Design for a Campus Hostel Building …

349

Table 4 Total energy consumed Parameters

CFL

No. of lamps

2

LED

Total load (W)

1

30

Energy consumed per room per annum (kWh)

227.2

96

Total no. of two bedrooms

252

Total energy consumed per annum (kWh)

57,254.4

24,192

Total cost (INR)

300,585

127,036

Fig. 13 Payback period of PSBC LED driver

a two-bedded room with the PSBC driver. It is also estimated that total cost to retrofit PSBC for 252 rooms costs is around INR 6.75 lakhs. Though the payback period of PSBC is higher as in Fig. 13, there is a considerable energy savings of INR. 1.73 lakhs per annum when compared with fluorescent lamps.

8 Conclusion The PSBC is designed for 6 and 24 W to replace the existing fluorescent lighting a two-bedded hostel building. The converters are experimentally verified for its high efficiency and output lumens. The efficiency of the PSBC is 5% higher than CBC. Illumination design calculations for PSBC and fluorescent-based two-bedded room are done. The ILER of PSBC driver is higher than fluorescent lamps. From the simulation results using DIALux software, the LOR of LED main and mirror light

350

R. Srimathi et al.

is one whereas for fluorescent main and mirror lights, the LOR is 0.81 and 0.4, respectively. Moreover, it is estimated that PSBC has good energy savings than the conventional AC and DC-CBC lighting. Therefore, PSBC is suggested as a better alternative in low-power lighting applications. Acknowledgements The authors are thankful to the VIT management for the support rendered to them.

References 1. Principi P, Fioretti R (2014) A comparative life cycle assessment of luminaires for general lighting for the office-compact fluorescent (CFL) vs light emitting diode (LED)—a case study. J Clean Prod 83(1):96–107 2. Newsham G, Arsenault C, Veitch J, Tosco AM, Duval C (2005) Task lighting effects on office worker satisfaction and performance, and energy efficiency. Leukos 1(4):7–26 3. Johnson RE (2010) Lighting device for accent lighting & methods of use thereof Google Patents, US Patent App. 12/614,935 4. Adkins E, Eapen S, Kaluwile F, Nair G, Modi V (2010) Off-grid energy services for the poor: Introducing LED lighting in the Millennium Villages Project in Malawi. Energy Policy 38(2):1087–1097 5. Li S, Lee ATL, Hui SY et al (2016) A plug-and-play ripple mitigation approach for DC-links in hybrid systems. In: 2016 IEEE applied power electronics conference and exposition (APEC), pp 169–176 (2016) 6. Dragicevic T, Vasquez JC, Guerrero JM, Skrlec, D (2014) Advanced LVDC electrical power architectures and microgrids: a step toward a new generation of power distribution networks. IEEE Electr Mag 2(1):54–65 7. Arias M, Vazquez A, Sebastian J (2012) An overview of the AC-DC and DC-DC converters for LED lighting applications. Automatika 53(2):156–172 8. Srimathi R, Hemamalini S (2018) Multi device integer frequency LED boost driver for outdoor lighting applications. J Adv Res Dyn Control Syst 10(1):234–244 9. Srimathi R, Hemamalini S (2018) LED boost driver topologies for low voltage DC distribution systems in smart structured buildings. Electr Power Compon Syst 46(10):1134–1146 10. Srimathi R, Hemamalini S (2019) Performance analysis of single-stage LED buck driver topologies for low-voltage dc distribution systems. IETE J Res 110:1–13 (2019) 11. Srimathi R, Sitoke S, Hemamalini S (2017) High efficiency buck led driver using sic. Energy Proc 117:224–235 12. Ravi D, Reddy BM, Shimi SL, Samuel P (2018) Bidirectional DC to DC converters: an overview of various topologies, switching schemes and control techniques. Int J Eng Tech 7(4.5):360–365 13. Guan X, Xu Z, Jia Q-S (2010) Energy-efficient buildings facilitated by microgrid. IEEE Trans Smart Grid 1(3):243–252 14. Chung MH, Rhee EK (2014) Potential opportunities for energy conservation in existing buildings on university campus: a field survey in Korea. Energy Build 78:176–182 15. Mangkuto RA (2016) Validation of DIALux 4.12 and DIALux evo 4.1 against the analytical test cases of CIE 171: 2006. Leukos 3(12):139–150 16. Omkar K, Karthikeyan KB, Srimathi R, Venkatesan N, Avital EJ, Samad A, Rhee SH (2019) A performance analysis of tidal turbine conversion system based on control strategies. Energy Proc 160:526–533

Lighting Design for a Campus Hostel Building …

351

17. Yu S, Tan Q, Evans M, Kyle P, Vu L, Patel PL (2017) Improving building energy efficiency in India: state-level analysis of building energy efficiency policies. Energy Policy 110:331–341 18. Ministry of Statistics and Programme Implementation. http://www.aeee.in/wp-content/upl oads/2017/05/ECBC-presentation-by-BEE.pdf

Electric Vehicles

A Novel Approach for Hobby Class Remotely Operated Vehicle T. M. Thamizh Thentral, T. V. Abhinav Viswanaath, S. Senthilnathan, and R. Bhargav

Abstract Nowadays, technology is becoming very advanced. Machines are capable of increasing efficiency, precision, speed and reducing the risk factor encountered by people. Our project deals with one such case concerning underwater exploration. Many people are finding it difficult to go underwater for taking measurements, exploring the sea bed and surveying before commissioning some mega off-shore projects due to factors such as strong currents and lack of knowledge about what is inside. The proposed system will go underwater and help us in getting a live stream of what is happening inside. It incorporates three-axis (x, y, z) movement and is provisioned with a light control to increase the visibility underwater. It comprises of ground station, communication link (tether) and the vehicle itself. It is a hobby class remotely operated vehicle implemented using Px4 as the central flight controller, Raspberry Pi as companion computer and QGroundControl as the ground station. Keywords QGroundControl · Px4 · Raspberry pi · Hobby class

1 Introduction The remotely operated vehicle will be referred as ROV here afterward. ROVs have become a crucial platform for doing scientific, industrial and civil operations in the marine environment [1, 2]. Compared to the use of divers, ROVs can go deeper, stay for a long time underwater and conduct complicated tasks consuming lesser amount of time [3]. Compared to automated underwater vehicle (AUV), ROVs are more suited for local, low-speed inspection and intervention tasks, particularly, those that involve human piloting or supervision as well as those requiring real-time streaming of captured data. An ROV is a tethered underwater maneuverable vehicle which finds use in various sectors. It can be classified into military, science, broadcast, hobby and many other classes. This work concentrates on hobby class ROV [1]. Hobby class ROVs are T. M. Thamizh Thentral (B) · T. V. Abhinav Viswanaath · S. Senthilnathan · R. Bhargav Department of EEE, SRM Institute of Science and Technology, Kattankulathur, Chennai, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_27

355

356

T. M. Thamizh Thentral et al.

generally built for experimental purposes. Several competitions have emanated to award most efficient ROVs such as MATE [4], Arab ROV competition, IET and UWM ROV competitions. The hobby class ROV generally possess a ground station which connects to a flight controller (which will be referred as FC hereon) underwater via a tether cable. Camera, lighting and thruster setup is connected to the FC for effective control of camera tilt, brightness and speed of thrusters, respectively. In this case, the FC is accompanied by Raspberry Pi which acts as a companion computer. The companion computer possesses a UNIX-based operating system called Raspbian which is IP configured to work with QGC at ground station. MAVLink protocol (will be discussed) is used for communication between the ground station and Raspberry Pi through a CAT5 tether cable. There are endless applications of ROV. Within the stipulated time, we could only make a completely functional hobby class ROV prototype. However, this can be upgraded to incorporate computer vision for making it applicable to a variety of industries starting from military and going till scientific experimentation such as analyzing behavior of a certain type of fish [3, 5, 6]. As the materials used are affordable, it can be easily accessed by any group of people and can be modified according to their needs.

2 System Design While designing the system, the following 10 principles as stated by Bohm and Jensen and used by the MATE center for its educational training have to be taken into account, “(1) water pressure is always there. (2) A structure keeps it all together. (3) Bigger is not necessarily better. (4) A submersible should float before it is ballasted. (5) The weights always end up below the floats. (6) Moving ballast weights causes tilt. (7) Any submersible has to be able to move. (8) A submersible needs some type of energy source. (9) You have to be able to control the power. (10) It is important to be able to navigate the submersible” [4].

2.1 Design Considerations Several of the design choices made for the ROV were based on cost-effectiveness, ease of use and ease of upgrade. According to our goal, the ROV should meet the following requirements: (1) The ROV should have six degrees of freedom, (2) it should protect all electronic components from water, (3) any slight movement should be observable by incorporating appropriate set of gyro and accelerometer sensors. (4) Control commands from the ground station located at the surface should be transmitted through a tether cable. Sensor and video stream should be transmitted back to the operation console by the same means and (5) it should be possible to add other modules to the frame for special applications.

A Novel Approach for Hobby Class Remotely Operated Vehicle

357

As water depth increases, the density and hence the water pressure increases. Therefore, the depth up to which the ROV can go is limited by the quality of waterproofing of the electronic components used in the watertight enclosure and the frame. It also depends on the torque output of the motors and the type of power supply. As discussed by Cook [7], the increase in battery capability has led to a stage where a few factors such as penetration distance and duration can be traded off to incorporate battery power, thus removing the need of powered fiber links. We have used a 5200 mAh lithium polymer battery with a rated voltage of 14.8 V which can withstand high load current and as this is a hobby class ROV, the frame material should be readily available and the chosen material should be corrosion resistant. Further, to implement the fourth principle stated by Bohm and Jensen, the material should possess the property of floating in water without adding any significant weight to the setup. PVC pipes were chosen to build the frame. As a very hard and permanent waterproofing was required on the joints and opening, it was done with epoxy for protection and M-seal for rigidity. As a high torque was needed to reduce the drag experienced due to high water density, BLDC motors were used for thrusters. Considering these factors, it would be able to go till a depth of 50–100 ft as typical Hobby class ROVs do. It can stay for 30 min which is limited only by the battery capacity and efficiency in control.

2.2 System Overview The overall system consists of the vehicle, the tether, the top-side ground station control and monitor as shown in Fig. 1. The QGroundControl software at ground station allows the user to view sensor data, video stream and to turn on and off components such as light and camera. The ground station sends commands to the vehicle via the tether, this command is passed from Raspberry Pi which is the companion computer to Pixhawk. The Pixhawk is the flight controller which controls the electronic speed controllers (ESCs) through PWM and in turn varies the motor speed. Additional LED light strips are used to assist camera focus. The 5MP H.264 Raspberry Pi camera captures and sends a live relay to the ground station via tether. Thus, the actions performed in the joystick connected to laptop results into vehicle motion as the joystick positions are mapped to the duty cycle of PWM sent by the flight controller.

3 Hardware Component The hardware can be divided into two systems, namely ROV frame with a watertight electronic housing and electronic component. The entire electronics is kept inside a 90 mm PVC pipe with acrylic windows for camera. Motor mounts are provided at appropriate places according to the standard 3 thruster setup.

358

T. M. Thamizh Thentral et al. Ground Station PC

Raspberry Pi

Pi Camera

Joystick control

Pixhawk Px4

ESC #1

BLDC Motor #1

ESC #2

BLDC Motor #2

ESC #3

BLDC Motor #3

Lighting

Underwater setup Unidirectional communication Bidirectional communication

Fig. 1 Hobby ROV block diagram

3.1 ROV Frame To meet our goal, the ROV frame should possess the following properties: (1) it should allow the ROV to freely rotate with six degrees of freedom, (2) it should allow easy upgrade by adding extra modules such as actuating arms, SONAR and other measurement modules [5]. (3) Materials used should be resistant to different conditions of water. (4) It should provide a supporting structure which can hold the thrusters at proper places. (5) The frame should provide an enclosure to hold all the electronic components which is not affected by disturbances in supporting structure. (6) Materials should be accessible and economical [8]. It is desired that the density of the ROV frame should be greater than that of water to make the ROV submerge. However, as PVC pipes float in water due to the air trapped inside, 8 holes are drilled at elbow joints such that water enters inside when ROV is submerged and trapped air goes out. Thus, the density and weight of the frame increases. This acts as a ballast. Air-filled water bottles are used as floats. For gluing the PVC pipes, PVC primer is used. The enclosure must be completely watertight as any leak can damage the electronics inside. For this reason, we have used 5-min epoxy which is readily available in the market. For giving an extra protection, M-seal is used on top of epoxy to hold the acrylic rigidly with the end cap. The length, breadth and height of the ROV are taken as 45, 33 and 25 cm, respectively. Further, the frame is constructed having 6 vertical and 8 horizontal PVC

A Novel Approach for Hobby Class Remotely Operated Vehicle

359

Fig. 2 a Top view, b Front view, c Isometric view

sections connected rigidly using 8 elbow and 18 tee joints ensuring there is no imbalance in inertia which may push the setup toward one side. The setup is constructed after drafting a rough CAD model as shown in Fig. 2a–c, respectively. The height of the left and right motors must match the vertical center of mass of the ROV. This makes the motors to give a level thrust to the ROV. If the motors are mounted too low or high, the resultant thrust may push one side of the ROV forward causing it to topple [9, 10]. 2.75 inches model marine propellers were attached to the motor shafts. The couplers to attach the propellers to the motors were made from a 2 mm ID thread which consist of a smaller shaft and a hole drilled in a bigger cap to fit the motor shaft. For attaching the top thruster, a rectangular 150 mm acrylic with two 3 mm holes were made using laser cutting, whereas the left and right side motor mounts are directly drilled to the PVC. On one end of the 90 mm electronic housing, a big threaded plug with a threaded cap (with washer) is used to make it easier to repair the components in case of faults. On the other end, a normal 90 mm end cap is used. A circular hole with 60 mm diameter is drilled from the center of the end cap and instead sealed with a transparent circular acrylic of same diameter done through laser cutting. The submersion of the motors in water does help to dissipate some heat and hence no protection as such is given for the BLDC motors, but the bearings need to be reoiled after each use to improve efficiency. The motor mounts were extended outward to give more room for the propellers to push the water and they are kept in such a way that the two motors in yaw axis are not in line with the motor in pitch axis in order to give it a lateral push.

3.2 Power Distribution The power flow diagram indicates the flow of power across all components as shown in Fig. 3. The power distribution starts from the battery and goes to the power module which consists of an LM317 voltage regulator along with 12, 5 and 3.3 V outputs for tapping. However, it is only used as a voltage distributor for tapping five 12 V

360

T. M. Thamizh Thentral et al. LiPo Battery (12v)

Power Module (12v)

ESC #1 (12v)

LED LIGHT STRIP (12v)

ESC #2 (12v)

BUCK CONVERTER (12v/4.8v)

ESC #3 (12v)

PIXHAWK (5v)

RASPBERRY PI (5v)

Fig. 3 Power flow

outputs and not used for directly taking the 5 V outputs required for Raspberry Pi. This is because, the current corresponding to the remaining 7 V have to be dropped as heat and tapping two of it makes things worse even though it will be operable. The heat loss is calculated using the Eqs. (1–4). Pin = Vin ∗ Iload

(1)

Pload = Vload ∗ Iload

(2)

Which implies, Pin = 12 V ∗ Iload

Which implies, Pload = 5 V ∗ Iload Vdrop = Vin − Vload

(3)

Preg = (Vin − Vload ) ∗ Iload Preg = (12 V − 5 V) ∗ Iload

(4)

Therefore,

where Pin V in I load V load V drop Preg

Input power in Watts Input voltage in V Load current in Amps Load voltage in V Voltage drop across regulator in V Power dissipated by regulator in Watts

A Novel Approach for Hobby Class Remotely Operated Vehicle

361

Thus, for a 120 mA load current drawn by Raspberry Pi and 60 mA for Pixhawk, we will have: Preg = 7 V ∗ .18 A = 1.26 W This 1.26 W will be dissipated as heat. Therefore, an external buck converter which is a switching regulator is used to share the load and results are very pleasing as the converter uses mosfet which is more efficient and emits less heat. Thus, the 3 ESCs, 1 light and 1 Buck converter are powered up with the 12 V. The 5 V output from the buck converter is given to Raspberry Pi and Pixhawk in parallel.

3.3 Flight Controller For controlling the speed of the motor, several techniques exist such as PI, simple FLC and Fuzzy-PI controller [11–13]. But, fuzzy controllers in general lacks realtime response and also requires more time and resources to run. Thus, in our work, we will be using a proportional integral derivative (PID) controller which is more robust to mismatches in tuning and it is also easier to implement and tune while taking less resources. The PID controller is integrated in flight controller called Pixhawk. Pixhawk is flight controller with an advanced auto-piloting capability designed by the “PX4 open-hardware project.” Processor and sensor technology are imported from ST microelectronics and real-time operating system are provided by NuttX for delivering efficiently while also having flexibility and reliability for controlling any remotely operated vehicle. It comes with an accelerometer and a compass on board which is calibrated before fight. Moreover, it allows interfacing many other sensors such as barometer, leak detector and GPS. It consists of 14 PWM/servo outputs in which 8 come with failsafe and manual override and 6 are auxiliaries which are compatible with higher power devices. It takes in 5 V supply and outputs a PWM pulse for controlling the ESCs. It also comes with low voltage indication and can be either plugged into ground station or can be interfaced with an intermediate companion computer to extend the functionality and enable tether communication. The PID loop helps to position the ROV with respect to the joystick command. Proportional controller serves the purpose of detecting the present position (displacement) and applying an equal amount of counterforce to bring the ROV to the new position. But proportional control by itself produces a wobbling effect as the counter control pulse may be high in such a way that it introduces oscillations and thus takes time to reach the mean position. Integral control captures the past position and gradually increases the control pulse until it reaches the mean position and thus very slow by itself. Derivative control predicts the future position based on current position and angular speed and makes sure that the current pulse is sent in such a way that it does not cause any disturbance in the future. Thus, a perfect combination of proportional, integral and derivative control as given in Fig. 4 is integrated into Pixhawk by tuning the Kp, Ki and Kd values to

362

T. M. Thamizh Thentral et al.

P

+

Error

I

-

+

+

Actuator

ROV

+

D

Feedback Fig. 4 PID loop

make the ROV act as per our needs [11]. The Pixhawk takes the feedback through the ESC and outputs an appropriate PWM pulse. In our setup, the first, second and third main channel outputs are connected to the ESCs corresponding to left, right and top thrusters, respectively. The speeds of motors are adjusted by varying the duty cycle of the PWM pulse. More ON time will result in a higher speed. The joystick positions are mapped to PWM outputs corresponding to the number and position of thrusters in the frame. The Pixhawk is loaded with ArduSub firmware through a 16 GB memory card which also gives the storage for storing videos.

3.4 Electronic Speed Controllers (ESCs) ESCs are used to control the speed of BLDC motors in the thruster assembly. The ESC takes in 12 V to power the motor and takes care of switching the phases of BLDC motor according to the PWM provided by Pixhawk. The speed is increased by increasing the switching frequency. Thus, more the duty cycle of the PWM, more will be the switching frequency of the ESC. An ESC consists of FET switching circuit which varies the ON time of FET to produce an increase of current in one phase [14]. The MOSFET switches at a rate of 2000 times per second. This technique however has a drawback in certain portions of joystick where the rest of the power from battery has to be released as heat from resistor. Thus, ample cooling should be provided. In our work, the water by itself acts as a cooling medium. As this is an Opto ESC, it supports an operating frequency of up to 600 Hz which is greater than that of Pixhawk which is 400 Hz and it also dissipates less amount of heat. It comes in different types: Brushed and brushless ESC, ESC with BEC and UBEC, ESC with BLHeli, simonk or kiss firmware. In our case, we have used 3 brushless Opto UBEC ESC with BLHeli S firmware (Fig. 5) as it is found to respond well to

A Novel Approach for Hobby Class Remotely Operated Vehicle

363

Fig. 5 Electronic speed controller

Pixhawk’s PWM signals. The ESCs are highly efficient, smooth and programmable as it hosts a BLHeli firmware. They have to be chosen with respect to your motor rating. In our work, we have used a 25 A motor and thus we have used a slightly higher 30 A ESC. An Arduino Nano ATMega328 board is made as a 4-way programming interface and used for programming the ESC through the servo pins. BLHeli suite is used for reading the setup and writing the new setup. The rotation mode is set to bidirectional and the PPM min throttle is set to 1100, PPM max throttle to 1900 and PPM midpoint to 1500. This setup is done because at zero throttle, the Pixhawk output will be at 1500 ms and if this comes in throttle range, it will cause undesirable effects.

3.5 ROV and Thruster Performance Calculations Certain parameters which are necessary for evaluating the efficiency of the system are maximum speed, maximum horsepower and time taken to stop ROV. Maximum speed is easily calculated by the distance traveled by the ROV in 1 s denoted by D1s. D1s = 0.6 m Therefore, the maximum speed, β max = 0.6 m/s. The maximum speed of the ROV can be improved by using a more powerful motor of lower KV as torque which is needed to overcome the drag is in lower KV motors. Batteries of higher capacity can be used and the contour of the propeller can be altered to improve hydrodynamics. Denoting time taken to cover one revolution along Yaw and Pitch axis as T y and T p, Ty = 8 s and Tp = 10.5 s

364

T. M. Thamizh Thentral et al.

Denoting angular speed along the Yaw and Pitch axis as N y and N p , respectively, Ty = 8 s and Tp = 10.5 s     Ny = 60/Ty ∗ 2π and Np = 60/Tp ∗ 2π

(5)

Substituting the values in Eq. (5), N y = 15π rad/min and N p = 11.43π rad/min. The angular speed along the Pitch axis can be increased by using 2 thrusters instead of 1 bidirectional thruster. max rpm = K v ∗ Vmax

(6)

Substituting K v = 800 kV in Eq. (6) as we have an 800 kV motor max rpm = 800 ∗ 12 = 9600 rpm where K v is expressed in RPM/V From Eq. (7) as proposed by “Hendershot and Miller,” the torque constant K q is given by, K q = 30/(π ∗ K v ) K q = 30/(3.14 ∗ 800) = 0.012 Nm

(7)

As we prefer motor efficiency over power, maximum horsepower loses its scope and maximum efficiency η comes into play, the general formula for efficiency as given in Eq. (8) can be rewritten as Eq. (9). η = Mechanical output power/electrical input power

(8)

η = (V − I ∗ Rm) ∗ (I − Io )/(V ∗ I ) Io = 0.9 A, I = 5.7 A, V = 9 V, Rm = 0.041 Ohm

(9)

Substituting the above values, we get the motor efficiency η = 0.82 or 82% where, V and I denote values of working voltage and current Rm Internal winding resistance in Ohms Io Current at no load. Current at maximum efficiency I max is needed to calculate torque and hence can be calculated using Eq. (10), Imax =

√ (V ∗ Io /Rm)

A Novel Approach for Hobby Class Remotely Operated Vehicle Table 1 Efficiency parameters

365

Maximum speed of ROV (m/s)

0.6 m/s

Maximum angular velocity (Yaw axis)

15π rad/min

Maximum angular velocity (Pitch axis)

11.67 rad/min

Time taken to stop ROV (from 0.6 to 0 m/s)

1.5 s

Torque at maximum efficiency (Nm)

0.1578 Nm

Maximum rpm of motor

9600 rpm

Motor efficiency

82%

Imax =

√ (9 ∗ 0.9/0.041) = 14.05 A

(10)

Torque at max efficiency can be given by Eq. (11) Q max = K q (Imax − Io ) Q max = 0.012 ∗ (14.05 − 0.9) = 0.1578 Nm

(11)

The parameters are tabulated in Table 1.

4 Software Component 4.1 Mavlink Protocol Mission air vehicle link---a lightweight protocol for messaging and communicating with drones between onboard ROV components and other ArduSub pilot softwares. MAVLink follows a hybrid “publish-subscribe and point-to-point” design pattern [11]. The sending of data stream is done as topics and configuration sub-protocols such as the “mission protocol” or “parameter protocol” are point-to-point with retransmission. The protocol is incorporated in autopilots such as Pixhawk [15]. The definition of messages is done through XML files also referred to as “dialect.” Most ground control stations implement a reference message set and is defined in common. Each message consists of 17 bytes having 6 bytes for the header, 9 bytes for the payload and 2 bytes for the checksum. The header contains information about the message such as vehicle type, flight controller used, system ID, message length, message type, and an incremental packet number. The payload contains the message data. The checksum is used to verify the integrity of the message and it is not altered during transmission. A maximum of 256 vehicles can be controlled using a ground station as it is only limited by number of vehicle IDs available in the protocol. XML message definitions are used by the MAVLink toolchain for generating MAVLink libraries specific to the supported programming languages. The generated libraries are used by ROVs, ground control stations and other such MAVLink systems to communicate with each other. The

366

T. M. Thamizh Thentral et al.

protocol is also dependent on the type of communication medium, which in our case is a CAT5 tether cable [16].

4.2 Ground Station QGroundControl is a mission planning software compatible with Pixhawk flight controller. It is used as a ground station for Rovers, Copters and Submersibles. It uses MAVLink protocol to communicate with the vehicle. It gives access to all the sensor data and gives the ability to steer the vehicle through a virtual joystick. It supports multiple flight stacks and also allows managing more than one vehicle. It gives us the live video stream and also an ability to take photos. It provides failsafe modes which are very useful for critical systems. In our work, we have used ArduSub firmware for Pixhawk which is flashed using this interface through a direct USB connection and can be updated anytime. This gives us the complete control over the ROV. Various Pixhawk parameters can be changed and it also gives a good amount of control over the auxiliary devices such as camera gimbal, lighting control, leak detectors and surge protectors. The tether connection is given from companion computer to the laptop and type of frame is selected which in our case is a 3-thruster simple ROV frame and the sensors are calibrated before the flight. Initially, individual motor test is performed and the corrections in rotation of thruster are done through software. The joystick is calibrated and the vehicle is armed for preparing to fly. The QGC is capable of communicating only with vehicles whose autopilots communicate through MAVLink protocol.

5 Integration and Test Results Thus, the electronic circuit is put into the watertight enclosure having the camera which is connected to Raspberry Pi to face the window. The battery is kept in the outermost space for ease of removal and charging. The watertight enclosure is band clamped and made height adjustable after which the thrusters are screwed in place using the motor mount. The lights are kept in front to assist camera focus in absence of daylight. The wires are given through a hole drilled in a threaded plug which couples the external motor and light wires with electronics in the enclosure (Fig. 6). During dry test, the motors were made to spin on their full speed without any load and the no load current and back EMF of the BLDC motor were calculated. Noise level and estimated duration of flight were checked. Operation of lights was also tested and a few short circuits were identified and rectified. During wet test, the setup was found to be floating. To make it submerge, holes were drilled at corner elbow joints of the frame which enabled water to go inside and thus act as a ballast. Also, it gave a clear passage for trapped air to escape out which was observed as bubbles (Fig. 7).

A Novel Approach for Hobby Class Remotely Operated Vehicle

367

Fig. 6 Dry run-motor test

Fig. 7 ROV in test tank

Now, as the setup went down, bottle floats were added to balance the ROV. Thus, making the ROV operable. Though the effect of drag added a significant load on the thrusters, on the flipside, noise levels were significantly damped and the direct contact of water with the motors assisted by being a lubricant. Now, the maneuverability of ROV was tested and the six degrees of freedom were tested out.

6 Future Scope This setup can be further upgraded to find use in diverse areas such as military surveillance, topography analysis before constructing off-shore plants. It can even be made to help the rural fishing communities in finding a rich location and also in drift netting [3]. It can even be used for educational purposes and scientific research on aquatic life [1]. In present situation, India is in lack of minerals in land. Whereas, we have abundance of minerals underwater in the Indian Ocean. In the future, we may even need to start consuming fossil fuel from oceans which is a hard truth. In

368

T. M. Thamizh Thentral et al.

such cases, this ROV can be used to detect the presence of minerals over an area [6]. Thus, the skeleton of the ROV which we have made can be upgraded to have computer vision to cater a variety of areas and it will considerably reduce the human risks and capital involved to perform such tasks.

References 1. Poore K, Kitts C, Wheat G, Kirkwood W (2016) A small scale ROV for shallow-water science operations. OCEANS 2016 MTS/IEEE Monterey 2. Capocci R, Dooly G, Omerdic E, Coleman J, Newe T, Toal D, Inspection-class remotely operated vehicles. Published in a review. J Marine Sci Eng 5:13. https://doi.org/10.3390/jms e5010013 3. Rubin S (2013) Mini-ROVs, going where no ROV has gone before. Published in: 2013 OCEANS, San Diego 4. Erica L (2009) Moulton: ROV in a Bag—an introduction to Remotely Operated Vehicles (ROVs) for the classroom. OCEANS, London 5. Bezanson L, Reed S, Martin EM, Vasquez J, Barbalata C (2017) Coupled control of a lightweight ROV and manipulator arm for intervention tasks. Published in: OCEANS 2017, Anchorage 6. Negahdaripour S, Xu X, Khamene A, A vision system for real-time positioning, navigation, and video mosaicing of sea floor imagery in the application of ROVs/AUVs. Published in: Proceedings fourth IEEE workshop on applications of computer vision. WACV’98 (Cat. No. 98EX201) 7. President MC, Crandle T, Cook G, Celkis E (2017) Tradeoffs between umbilical and battery power in ROV performance. Published in: OCEANS 2017, Anchorage 8. Marzbanrad A, Sharafi J, Eghtesad M, Kamali K-S (2011) Design, construction and control of a remotely operated vehicle (ROV). Published in ASME 2011 international mechanical engineering congress and exposition 9. Guan Z, Zhang D, Lin M, Li J (2018) School of construction machinery. Shandong Jiaotong University, Jinan, China: “Mechanical analysis of remotely operated vehicle”, Published in: 2018 4th international conference on control, automation and robotics (ICCAR) 10. Sahili J, El-Hadi Hamoud A, Jammoul A (2018) ROV design optimization: effect on stability and drag force. Published in: 2018 6th RSI international conference on robotics and mechatronics (IcRoM) 11. Khanke PK, Jain SD (2015) Comparative analysis of speed control of BLDC motor using PI, simple FLC and Fuzzy—PI controller. Published in: 2015 international conference on energy systems and applications 12. Hu Q, Lu Z, Qian Z (2007) Research on a novel close-loop speed control technique of brushless DC motor. Published in: 2007 IEEE power electronics specialists conference 13. Santhosh Kumar G, Arockia Edwin Xavier S (2015) Brushless dc motor speed control using microcontroller. Published in Troindia 2394-0697, vol 2, no 2 14. Jun B-S, Kook Y-S, Park J-S, Won C-Y (2018) A development of electronic speed control (ESC) for PMSMs driving used in drone. In: Conference: 2018 IEEE international power electronics and application conference and exposition (PEAC) 15. Atoev S, Kwon K-R, Lee S-H, Moon K-S (2017) Data analysis of the MAVLink communication protocol. Published in: 2017 international conference on information science and communications technologies (ICISCT) 16. Li Q, Xu H, Zhang Q, Wang X, Li Z (2008) Dynamics simulation of remotely operated vehiclefiber optic micro cable system. Published in: 2008 7th world congress on intelligent control and automation

Simulation and Analysis of Interleaved Buck DC-DC Converter for EV Charging K. Murugappan, R. Seyezhai, G. Kishor Sabarish, N. Kaashyap, and J. Jason Ranjit

Abstract Electric vehicles use electric power as the driving force instead of the fossil fuels for propelling vehicles. The main objective of introducing electric vehicles is to reduce pollution caused by burning fossil fuels. One of the problems in electric vehicles is charging the battery effectively to increase the lifetime of the electric vehicle (EV) battery. When conventional buck converter is used in EV charger, the output voltage and current contain ripples which reduce the life of the battery. To overcome this issue, a multi-phase interleaved step-down converter is employed which results in reduced ripples both at the input and output. In this paper, a comparative study between the topologies of the conventional and the two-phase interleaved buck converter is analyzed. The simulation studies are done using MATLAB-Simulink. The performance of interleaved step-down converter is compared in view point of load voltage and load current ripple, losses and efficiency with the classical stepdown converter. The comparisons show that the concept of interleaving reduces the ripple and increases the efficiency of the EV charger Keywords Interleaved buck converter · Voltage and current ripple · Efficiency

1 Introduction The vehicles in operation now use fossil fuels such as petrol, diesel which leads to increase in pollution levels. Fossil fuels are depleting at a faster rate, so it is necessary to find other choice of energy source to power up the vehicles. The concept of electric vehicles was introduced to overcome these shortcomings where electricity is used as fuel, generated not only from non-renewable resources but also from other renewable sources such as solar energy, wind energy, etc. Another advantage of implementing K. Murugappan (B) · G. Kishor Sabarish · N. Kaashyap · J. Jason Ranjit Department of EEE, SSN College of Engineering, Kalavakkam, Chennai 603110, India e-mail: [email protected] R. Seyezhai (B) · G. Kishor Sabarish · N. Kaashyap · J. Jason Ranjit Renewable Energy Conversion Laboratory, Department of EEE, SSN College of Engineering, Kalavakkam, Chennai 603110, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_28

369

370

K. Murugappan et al.

electric vehicles is that pollution is reduced as electric motors employed in these vehicles do not release pollutants. An important aspect involved in the concept of electric vehicles is the charging of the battery that supplies power to the motor and other devices in vehicles. An EV charger comprises of a conventional buck converter which produces ripples in the output voltage waveform that will damage the cells of the battery and might affect the state of charge and depth of discharge characteristics of the battery. Consequently, range of the EV is also reduced. In the past few decades, several topologies of buck converters were investigated and analyzed for better performance in terms of output and input parameters. One of the topologies uses the concept of interleaving to reduce the output current ripple [1]. The switching losses in the Interleaved Buck Converter were found to be lower than that of the conventional topology [1, 2]. Moreover, interleaving results in high power density and high efficiency [3]. The Interleaved Buck Converter has application in dual voltage power for passenger cars [4]. There exists works which are primarily focused on the distributed approach for interleaving power converters [5]. This paper discusses about the two-phase interleaved buck converter for EV battery charging applications. Its circuit operation, design equations and simulation studies are investigated and its performance is compared with the classical buck converter. From the simulation results, it is found that interleaved buck converter provides a better performance against its counterpart. The paper is planned as follows: Sect. 1 provides the need for interleaving concept; Sect. 2 reveals the basic buck converter operation and the simulation results. Section 4 focuses on the multi-phase interleaved step-down converter and its design equations. Section 5 gives a comparison assessment between the two topologies of the step-down converter and Sect. 6 concludes the paper.

2 Basic DC-DC Buck Converter The step-down DC–DC converter, referred as a buck converter, is shown in Fig. 1.

Fig. 1 Circuit diagram of DC-DC buck converter

Simulation and Analysis of Interleaved Buck …

371

In this circuit, the device is connected in series with the supply followed by freewheeling diode, filter inductor, capacitor and load R. When the switch S is triggered to conduction state, the diode D is reverse biased. When the switch S is switched off, the freewheeling diode comes into action providing a continuous current in the inductor [6]. The step-down converter operates in two modes: continuous current conduction and discontinuous conduction modes. In the continuous conduction mode (CCM) the current through the inductor is always more than zero. For CCM the switching frequency of the switch has to be higher and the value of inductance is higher too. In the discontinuous conduction mode (DCM), the inductor current reaches zero before the next switching cycle. For DCM, the switching frequency of the switch has to be lower and the value of inductance has lower too. For this work, buck converter operating in CCM is chosen. During the on state, the inductor is charged and the flow of current path is via the switch and to the resistive load. In the off state, the inductor is discharged and the energy stored in the inductor is dissipated in the load resistor as the current flows through the resistor and the freewheeling diode and back to the inductor. The conversion gain of the converter is given by Vo = Vs (K )

(1)

where K is the duty ratio of the MOSFET used. The duty cycle is given by K =

ton ton + toff

(2)

The design equation of buck convertor involves the calculation of the values of R, L and C for the desired load voltage, current and its ripple. The assumptions involved are: Load voltage V o = 50 V, load current I o = 3 A, output load voltage ripple (V ) = 1.1% of output voltage = 0.55 V, current ripple (I) = 40% of output voltage = 1.2 A, duty ratio (K) = 50%, switching frequency (f ) = 20 kHz. R=

Vo Io

(3)

L = Vs (1 − K )/If

(4)

C = I /8 f Vc

(5)

where V o is output voltage, I o is output current, R is value of load resistance, L and C are inductance and capacitance values of the LC filter, respectively, V s is source voltage, K is the duty ratio of MOSFET switch, I is the current ripple of output current, V is the voltage ripple of output voltage and f is the switching

372

K. Murugappan et al.

Fig. 2 Simulink diagram of step-down converter

frequency of switch. The DC-DC buck converter is simulated using MATLAB. The MATLAB/Simulink circuit is depicted in Fig. 2. Simulation parameters of the step-down converter are shown in Table 1. Figure 3 shows that the load current of step-down converter that has the high ripple for the given time period. Figure 4 shows the output voltage versus time waveform of classical step-down converter with ripple. Table 1 Parameters of step-down converter Parameter

Values

Vs

100 V

R

16.6667 

L

1.0417 mH

C

13.63 µF

Vo

50 V

Io

3A

Fig. 3 Load current of step-down converter

Simulation and Analysis of Interleaved Buck …

373

Fig. 4 Load voltage of step-down converter

3 DC-DC Interleaved Buck Converter The schematic diagram of two-phase interleaved step-down converter is shown in Fig. 5. It consists of two step-down converter stages in parallel connection and the gating signals for the two MOSFET switches are displaced by 180o , for a duty cycle of 50% to minimize the ripple to a maximum extent. The inductor currents I L1 and I L2 add up to give the final output current I o . Interleaving or multi-phasing is a concept used in converters to reduce the size of filters for equivalent ripple current or reduce the ripple by using filter of sizes used in conventional converters [7, 8]. The above circuit operates in two modes depending on the time period of operation (i) Mode 1: 0° ≤ firing angle ≤ 180°

Fig. 5 Schematic diagram of two phase interleaved step-down converter

374

K. Murugappan et al.

Fig. 6 Simulink representation of two phase interleaved step-down converter

Switch S 1 is in conduction state. Diode D2 is in on state. The current in the inductor L 1 increases linearly and inductor L 2 discharges whose energy is transferred to the load via the free-wheeling diode D2 . These operations occur simultaneously maintaining the output current value with most minimal ripples. (ii) Mode 2: 180° ≤ firing angle ≤ 360° Switch S 2 is triggered and it is forward biased. Diode D1 comes to on state. The inductor L 2 starts charging and the current through it increases linearly and the inductor L 1 discharges supplying the energy to the load. These operations occur simultaneously maintaining the output current with most minimal ripples. As the inductor currents vary inversely and as they are connected in parallel to form the output current the value of the output current is almost a constant. The design equation of buck convertor involves the calculation of the values of resistor, inductor and capacitor for the desired load voltage, current and its ripples. The assumptions involved are: Load voltage V o = 50 V, load current I o = 3 A, load voltage ripple (V ) = 1.1% of output voltage = 0.55 V, load current ripple (I) = 40% of output voltage = 1.2 A, duty ratio (K) = 50%, switching frequency (f ) = 20 kHz. The components are designed using Eqs. (1)–(5). The two-phase interleaved step-down converter is simulated using MATLAB and is shown in Fig. 6.

4 Design Steps The design of interleaved buck converter [9, 10] involves the following requirements.

Simulation and Analysis of Interleaved Buck …

375

1. Selection of number of phases The number of phases is selected as two for this work, for reducing the weight and size of the converter and construction will be simple as fewer components are involved. The choice of the number of phases is important since it decides the load voltage and input current ripple and as it increases, the ripple gets reduced drastically but more complexity arises in the implementation of the converter. 2. Selection of duty ratio The duty ratio is correlated to the number of phases and the phase number is two, the preferred duty ratio is 0.5 as the ripple gets nullified for this value. 3. Selection of switches The switches required for designing the converter are same as the number of phases; hence, two MOSFET switches are selected for the converter and these switches are triggered with 180 degree phase shift to achieve the interleaving operation [11, 12] but switching frequency remains the same. 4. Selection of inductors and capacitor The number of inductors is same as the number of phases; therefore, two inductors are chosen and one output capacitor with reduced ripple. Based on the above guidelines, the optimum design methodology is carried out for interleaved converter and the simulation parameters of the interleaved step-down converter are same as that of the conventional step-down converter (Table 1). Figure 7, 8 portrays the output voltage V o has less ripple when compared to conventional step-down converter’s output voltage waveform. Figure 9 shows that the two inductor currents I L1 , I L2 which is phase shifted due to interleaving concept with very less ripples. Figures 10, 11 and 12 represents the voltage stress of S1 and S2, current stress of S1 and S2, Current Stress of Diodes D1 and D2 respectively.

Fig. 7 Load voltage waveform of interleaved step-down converter

376

Fig. 8 Load voltage ripple of interleaved step-down converter

Fig. 9 I L1 , I L2 and I o waveforms of interleaved step-down converter

K. Murugappan et al.

Simulation and Analysis of Interleaved Buck …

Fig. 10 Voltage stress of S 1 and S 2

Fig. 11 Current stress of S 1 and S 2

377

378

K. Murugappan et al.

Fig. 12 Current stress of D1 and D2

5 Comparison of Topologies The analysis of buck and interleaved buck converter topologies is carried out and the following parameters are compared: (i) Voltage ripple: The voltage ripple in the conventional buck converter was found to be in the order of 10−2 V, whereas in the two-phase interleaved step-down converter, the load voltage ripple was found to be in the order of 10−6 V which is very less compared to the conventional buck converter. Therefore, by applying the principle of interleaving, the voltage ripple can be reduced which in turn increases the life cycle of the battery used in electric vehicles. (ii) Current ripple: The current ripple in the conventional buck converter was found out to be 1.196 A for the above specified design values which is quite high when compared to the current ripple in the interleaved step-down converter which was found out to be in the order of 10−6 A. If the current ripple is less as in the case of interleaved buck converter, the electric vehicles battery gets charged efficiently.

Simulation and Analysis of Interleaved Buck …

379

Table 2 Functional parameters Topology

Voltage ripple (V)

Conventional DC-DC buck converter 1.084 × 10−2 Two phase interleaved DC-DC buck converter

6.909 ×

10−6

Current ripple (A)

Efficiency (%)

1.196

98.99

6.521 ×

10−6

99.13

Table 3 Computation of losses Topology

MOSFET losses

Diode losses (W)

Total losses (W)

Efficiency (%)

0.0193

1.03862

1.53

98.99

0.02

1.0423

1.32

99.13

Switching losses (W)

Conduction losses (W)

Conventional DC-DC buck converter

0.4717

Two phase interleaved DC-DC buck converter

0.258

(iii) Losses: Losses play a major part in any topology that is employed and the losses decide the efficiency of the topology. So various losses such as MOSFET’s switching and conduction losses and diode’s conduction losses for both the topologies were computed using their respective formulae and it was found that the losses were higher for the conventional buck converter than the interleaved step-down converter though the number of devices used in interleaved converter are more. (iv) Efficiency: The efficiency of the topology is a very important parameter. The efficiencies of both topologies were computed and it was found out that the efficiency of interleaved step-down converter is higher than the conventional buck converter. The result of the comparison of topologies is listed in Table 2. From Tables 2 and 3, it is inferred that the interleaved step-down converter delivers a reduced load voltage ripple and current ripple with improved efficiency compared to the basic buck topology.

6 Conclusion Electric vehicles are of utmost importance for development in this day and age, and in order to make the operation of such electric vehicles economic, it is important to reduce the frequency of replacement of the battery which supplies power to the

380

K. Murugappan et al.

electric vehicle. A simulation and a comparative study of the topologies were done to provide further insight. The input current in case of conventional step-down converter is discontinuous causing ripple in the source side, whereas in case of interleaved stepdown converter, the input current is continuous and has reduced source side ripple. This paper proposed an Interleaved Buck Converter to reduce the voltage ripple and output current ripple considerably so as not to damage the battery while charging it, thereby increasing the life of the battery. And from the results, it is concluded that the interleaved step-down converter is the better choice as a charging topology as it has less ripple in the load current.

References 1. Esteki M, Poorali B, Adib E, Farzanehfard H, Interleaved buck converter with continuous input current, extremely low output current ripple, low switching losses, and improved step-down conversion ratio. IEEE transactions on industrial electronics 2. Lee O, Cho SY, Moon GW, Interleaved buck converter having low switching losses and improved step-down conversion ratio. In: IEEE transactions on power electronics 3. Oliver JA, Zumel P, García O, Cobos JA, Uceda J (2004) Passive component analysis in interleaved buck converters. In: 19th Annual IEEE applied power electronics conference and exposition. APEC’04 4. Shrud MA, Kharaz AH, Ashur AS, Faris A, Benamar M (2010) Analysis and simulation of automotive interleaved buck converter, World Academy of Science, Engineering and Technology, International Journal of Electrical and Computer Engineering vol 4, no 3 5. Garinto D (2006) A novel multiphase multi-interleaving buck converters for future microprocessors. In: 12th international power electronics and motion control conference, PEMC 2006, pp 82–87, Aug–Sep 2006 6. Hart DW (2011) Power electronics. Tata McGraw Hills, pp 237–238 7. Gallagher J (2006) Coupled inductors improve multiphase buck efficiency, power electronics technology, pp 36–42 8. Zumel P, Fernández C, de Castro A, Garcia O (2006) Efficiency improvements in multiphase converter by changing dynamically the number of phases. IEEE power electronics specialists conference 2006, June 2006 9. Gerber M, Ferreira JA, Hofsaer W, Seliger N (2004) Interleaving optimization in synchronous rectified DCIDC converters. In Proceedings IEEE power electron. Spec. Conf (PESC’04), pp 4655–4661 (2004) 10. Chen YM, Teseng SY, Tsai CT, Wu TF (2004) Interleaved buck converters with a singlecapacitor turn-off snubber. IEEE Trans Aerosp Electronic Syst 40(3):954–967 11. Perreault DJ, Kassakian JG (1997) Distributed interleaving of paralleled power converters. In: IEEE transactions on circuits and systems—I: fundamental theory and applications, vol 44, No 8, August 1997 12. Jakobsen LT, Andersen MAE (2007) Two-phase interleaved buck converter with a new digital self-oscillating modulator. In: 7th international conference on power electronics, October 2007

Design and Real-Time Implementation of Economical Solar Car for Commercial Applications Nabeel Najeeb Kassim, Sohail Akthar Khan, Kumar Devasish, A. Chitra, and Ryan Sujith

Abstract Our paper proposes an expedient design of a solar-powered hybrid electric vehicle for automation. An electric vehicle propels via a battery which has been stimulated by an external electric power supply. All contemporary electric vehicles proffer a drive on an extraneous AC power endowed motor. Contemplating those traditional setups where an inverter set is obligatory to be connected with the battery which in turn transforms the AC power to DC power. This alteration eventually results in considerable losses, besides the cost of maintenance of the AC system is steep. The proposed topology possesses the sovereign solar/electric power generation system mounted on the vehicle to energize the battery during all durations. The foremost functionality includes multi-charging via which vehicle can boost charge through both solar and electric power. A suitably coordinated network of battery, motor, battery management system along with maximum power point tracker which yields best results. Being cost-efficient, it serves the purpose of automation for a commoner. Keywords Solar module · Batteries · BLDC motor · Charge controller · Maximum power point tracker · Battery management system · Buck converter

1 Introduction Fossil fuels have become an expensive commodity to be processed and used. The exponential growth in the demand for vehicles that use internal combustion engines is the main reasons that have accelerated the extraction of these non-renewable resources. The most adverse effect of these fuels is the greenhouse effect caused by the combustion of these fuels where a large number of greenhouse gases are formed. Solar vehicles employ photovoltaic cells to convert sunlight directly into N. N. Kassim · S. A. Khan · K. Devasish · A. Chitra (B) · R. Sujith (B) School of Electrical and Electronic Engineering, Vellore Institute of Technology, Vellore 632014, India e-mail: [email protected] R. Sujith e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_29

381

382

N. N. Kassim et al.

electricity to supply the loads. Recent surveys have reported that fossil fuels are depleting at an alarming rate that has exponentially increased over the last couple of decades. It has now become the need of the hour to explore alternative energy resources. The Solar Electric Powered Hybrid Vehicle contains a solar panel, brushless DC motor, charge controller, batteries, and buck converter unit. Diode rectifier can be used to charge the batteries in normal AC supply during sunless conditions. The photovoltaic array has a particular operating point that can supply the maximum power to the load which is generally called maximum power point. To boost the efficiency of the PV system, the MPP has to be tracked and followed by regulating the PV panel to operate at MPP operating voltage point, thus optimizing the production of the electricity. One of the most important points in the construction of the vehicle is that it is closely related to the chassis design, to achieve a structural optimized work. The design was conceived from point of view of high efficiency, lightweight, and stable transport with reduced costs and zero-emission in its operation and as a source of energy, solar energy is being used to produce electricity with the help of this technology. We aim to design a solar energy powered vehicle in our paper. To power our vehicle in the absence of sunlight, an alternative power source was designed to run the vehicle. The vehicle combines the use of electric energy from the three different sources (a) Photovoltaic solar energy (b) Rectified power supply (charging) (c) Batteries.

2 Proposed Model The proposed SEPHV being cost-effective and eco-friendly serves as an effective means of automation for the future. The material used for making the chassis is light weighted, strong, and cheaper as compared to aluminium which is being generally used to make vehicles which in turn adds to the sets of benefits being provided by SEPHV. The solar cells used in the vehicle are monocrystalline solar cells which is the most efficient type of solar cell and is not as costly. The BLDC motor used is powerful, produces high torque, and requires less maintenance, and lightweight. The Li-ion batteries employed have less weight and are maintenance-free. These all factors and the precisely calculated parameters for battery, motor, solar panels, transmission systems, and vehicle dynamics made the proposed SEPHV lightweight, costeffective, and efficient. Figure 1 depicts the proposed model of the solar electric hybrid vehicle. The normal electric vehicle requires charging from household electricity. On the run, an electric vehicle cannot be charged while running while a SEPHV charges while travelling by a solar panel, this increases the travelling distance of the SEPHV as compared to a normal electric vehicle. When the car is parked, a SEPHV is being charged continuously saving time and money needs charging only during the night

Design and Real-Time Implementation of Economical …

383

Fig. 1 Proposed model of the solar electric hybrid vehicle

when the solar power is not available. This makes the vehicle more efficient and it travels greater distance per charge ideal for city use. On the comparison between a normal IC engine vehicle, an electric vehicle and a SEPHV based on pollution, a normal IC engine vehicle causes the most pollution releasing pollutant like carbon monoxide, carbon dioxide, water vapour, oxides of nitrogen, oxides of sulphur, etc. Comparing a normal electric vehicle and a SEPHV, the SEPHV cause the least amount of pollution, as stated above, a SEPHV needs less electricity charging as compared to a normal electric vehicle. Because 65% of the country’s electricity is generated by thermal power plants that involve the burning of fossil fuels which also contributes to heavy pollution. A SEPHV helps to reduce the dependence on electricity generated in power plants.

2.1 Chassis The chassis is made of Chromoly AISI (American Iron and steel institute) 4130 normalized alloy steel. Chromoly is an alloy of Chromium and Molybdenum. Chromoly is stronger than traditional 1020 steel [1]. It is easily formable, weldable and while welding, it does not lose its properties [2]. It has high weight-to-strength ratio compared to mild steel. Moreover, since we use long member, buckling is tackled by using Chromoly. The chassis was designed with prime focus on passenger’s safety. Many iterations were done so that in impact analysis, the force can be distributed uniformly to minimize the deflection in chassis members. The total weight of the chassis is only 34 kg. Figure 2 depicts the modelled chassis.

2.2 Brushless DC Motor and Its Characteristics The conventional DC motor has the advantage of easier speed control but has the disadvantage of wear and tear of commutators and electric sparks between brushes

384

N. N. Kassim et al.

Fig. 2 Chassis models

and commutators. These disadvantages can be overcome with BLDC motor because there are no brushes and commutators in this motor [3, 4]. The BLDC motor uses an inverter along with rotor position detection to control the armature currents which in turn control the speed of the motor. The maintenance is reduced in the motor as it is electronically commutated [5]. The commutation instants are found with the rotor position detection at each instant. The permanent magnet BLDC motor incorporates the advantages of higher efficiency, high reliability, ruggedness, and easy controllability, the capability to operate successfully at low voltage and excellent performance over a wide range of speed. The motor has low rotor inertia, allowing it to accelerate, decelerate, and reverse direction quickly. The BLDC motor has many advantages over brushed DC motors and induction motors, like better speed--torque characteristics, high dynamic response, high efficiency and reliability, long operating life, noiseless operation, higher speed ranges, and reduction of electromagnetic interference (EMI) [4, 6]. The specifications of the BLDC motor used in vehicle paper are as given in Table 1.

2.2.1

Motor Dynamics

A 3KW BLDC motor gives us a starting torque of 33 N-m. So, if we use a transmission ratio of 4.5:1, then the required starting torque required by the vehicle will be fulfilled. As we know, torque is inversely proportional to angular speed. Hence, the angular

Design and Real-Time Implementation of Economical … Table 1 Specifications of BLDC motor

385

S. No.

Specification

Range

1

Power

3000 W ± 5%

2

Rated voltage

48 ± 4 V

3

Rated current

62 ± 5 A

4

Rated speed

3000 RPM ± 4.5%

5

Rated continuous torque

14 N-m ± 2%

6

Peak torque

33 N-m ± 2%

speed will get decreased by a factor of 4.5. Weight of the vehicle = 310 kg Acceleration of the vehicle = 0.6 m/s2 Starting Torque for vehicle = 142 N-m Motor Starting Torque = 33 N-m Transmission Ratio = 4.5 : 1 Motor RPM = 2250 RPM Wheel RPM = Motor RPM/ Transmission Ratio = 2250/4.5 = 500 RPM Angular velocity of wheel (ω) = (2D ∗ 500)/60 = 52.33 rad/s Radius of the wheel(R) = 10 inch = 0.2413 m Speed of Vehicle = ω ∗ R = 52.33 ∗ 0.2413 = 12.62 m/s Speed of Vehicle(Km/h) = 12.62 ∗ 3600/1000 = 45.4 km/h. This 3KW BLDC motor runs the vehicle at a maximum speed of 45 km/h with an acceleration of 0.6 m/s2 .

2.3 Charge Controller and Battery A Lithium-ion battery is better than a lead acid battery, due to its higher greater energy density [7, 8]. In electronic equipment like mobile phones, laptops, etc., the need to operate for longer durations between charges while still consuming more power, there is always a need for batteries with a much higher energy density. The issue with batteries and cells is that they lose their charge over time. This self-discharge can be a major issue. One of the advantages of lithium-ion cells is that their rate of self-discharge is lower than that of other rechargeable cells such as Nickel Cadmium (Ni-Cad), Nickel Metal Hydride (NiMH) etc. [10, 11]. A distinct advantage is that lithium-ion cells and batteries need not be primed before their first charge. Another major advantage is that they do not require regular maintenance to ensure their

386

N. N. Kassim et al.

performance. Ni-Cad cells required a periodic discharge to ensure that they did not exhibit the memory effect. Lithium-ion batteries are nearly 100% efficient in both charge and discharge, allowing for the same amp hours both in and out. Lead-acid battery’s inefficiency leads to a loss of 15 amps while charging and rapid discharging drops voltage quickly and reduces the battery’s capacity [13]. There are multiple chemistries of lithium-ion cells available in the market. Certain types of lithium-ion batteries provide a high current density which makes them ideal for consumer mobile electronic equipment. Whereas the others provide much higher current levels and are ideal for power tools and electric vehicles [14, 15]. Considering safety and cost factors sealed lead acid batteries are preferred over Li-ion batteries, as Li-ion batteries cost nearly thrice as much as sealed lead-acid batteries. But when designing a highly efficient electric solar hybrid vehicle weight is a crucial factor. Sealed lead-acid batteries weigh nearly 380% more than the Liion batteries. Also, when designing the SEHPV, we have considered reliability and minimum maintenance as important aspects of a cost-efficient vehicle. The Li-ion batteries again prove to be a better option with greater shelf life, greater depth of discharge, lower self-discharge, and higher battery cycle. These factors offset the higher cost of Li-ion batteries. These factors make Li-ion batteries the better option when compared with other battery technologies.

2.3.1

Battery Design

Each Li-ion cell provides a 4.2 V. Such cells are arranged into a stack that is electrically insulated and fire-resistant. A stack of these Li-ion cells provides an overall voltage of 48.1 V and 130 Ah. Fourteen stacks of the same specification are connected in to meet the required demands. Battery specifications are given in Table 2. Due to the ebullient nature of the Li-ion batteries, we have incorporated various protection technologies. The entire stack setup is again insulated use SE-2 glass fiber material. This makes the setup highly resistant to shock and fire. Table 2 Battery specification

S. No.

Specification

Range

1

Voltage

48 ± 2 V

2

Peak current

140 ± 10 A

3

C rating

1C

4

Dimension

500 mm (l) * 300 mm (b) * 200 mm (h)

5

Continuous current

72 ± 5 A

Design and Real-Time Implementation of Economical …

2.3.2

387

Battery Setup

To power the 48 V BLDC motor for a minimum of 1.5 h at, we have chosen a 48 V 130Ah Li-ion battery pack. The losses such as friction, drag, and other mechanical losses are accounted. Considering that the losses use up at least 30 min of the run time available, we get a final output of 1.5 h run time at the proposed maximum speed of 44 km/h. The calculations leading to the result are as shown below: Rated Voltage = 48 V Required Ah = Continuous load current ∗ time continuous load current (Motor) = 62 Amps Time = 2 h(1.5 h + Losses) Required Ah = 62 Amps ∗ 2 h = 124 Ah Therefore, Minimum Capacity of Battery pack = 124 Ah Approx. Charging time at 20 A(Solar Charger) = 130/20 = 6.5 h. Ah of the battery pack is 130 Ah. This battery can run the vehicle for 1.5 h at a speed of 45 km/h continuously. A battery management system has been integrated into the battery circuit for safety and monitoring purposes. The specification of the BMS is as given in Table 3. Battery management system (BMS) limits the electric current supplied to or drawn from the battery pack. The primary function of the charge controller and BMS is to prevent overcharging and deep-discharging of the battery pack. The charge controller also consists of the circuit that controls the depth of discharge (DOD) of the battery.

2.4 Solar Panel Solar cells are solid-state semiconductor devices which convert light energy directly into electrical energy. A solar cell contains a low voltage typically about 0.45 V per cell; cells are connected in series to increase voltage. Figure 3 shows the equivalent circuit of a solar cell. The model of the solar cell can be categorized as P-N semiconductor junction when exposed to light; the DC is generated and the generated current depends on the solar irradiance, temperature, and load current. Figure 4shows the ideal current–voltage characteristics of a solar cell. Two types of solar cells are used, monocrystalline cells and back-contact semi-flexible monocrystalline cells [13, 17]. Table 4 depicts the specification of the solar module used in our vehicle.

388 Table 3 Battery management system specifications specifications

N. N. Kassim et al. S. No. Parameter

Description/rating

1

Cell technology supported

Lithium ion

2

Number of cells in series

13

3

Nominal voltage

48.1 V

4

Full charge voltage

54.6 ± 0.05 V

5

Single balance voltage

4.2 ± 0.025 V

6

Single balance current

42 ± 5 mA

7

Self-consumption current (single) ≤ 25uA

8

Continuous charge current

20 A

9

Continuous discharge current

20 A

10

Over current test voltage

200 ± 25 mV

11

Over current test current

150 ± 20 A

12

Working temperature

−40 to + 85 °C

13

Storage temperature

−40 to + 125 °C

14

Communication port

NA

15

CAN supported

No

Fig. 3 Circuit of photovoltaic cell

Considering design flexibility and cost, we bought the solar cells and have manufactured the panel according to the vehicle design. The non-flexible and flexibles cells are used according to the areas where the panel is placed.

Design and Real-Time Implementation of Economical …

389

Fig. 4 Ideal photovoltaic cell curve

Table 4 Specifications complete of solar module

2.4.1

S. No.

Specification

Range

1

Maximum power (Pmax )

764.4 W ± 5%

2

Maximum voltage (V pm )

48 ± 2 V

3

Maximum current (I pm )

15.9 A ± 3%

4

Open circuit voltage (V oc )

50.83 V ± 2%

Solar Panel Manufacturing

We have used two types of solar cells, monocrystalline cells (42 cells) and backcontact semi-flexible monocrystalline cells (168 cells). Now the voltage of a normal monocrystalline cell is 0.6 V (Rated) with the current of 7 Amperes (rated) and that of a back-contact semi-flexible monocrystalline cell is 0.5 V along with 7 A current (rated). So total voltage output = (0.6 ∗ 42) + (0.5 ∗ 168) = 25.2 + 84 = 109.2 V This is more than the required voltage of 48 V of the tractive system.

So according to power conservation, P = V ∗ I Total power = 109.2 ∗ 7 = 764.4 W Also, the output current.I , 109.2 ∗ 7 = 48 ∗ I I = 15.9 A The solar panels designed along with each of their specifications are given in Table 5.

390

N. N. Kassim et al.

Table 5 Panel-wise specifications of solar module (All measurements are in inches) Parameters

Panel 1

Panel 2

Panel 3

Panel 4

Panel 5

Location

Front

Top

Back

Front sides

Back sides

Dimension

43 × 38

40 × 20

40 × 55

5 × 30 (both sides)

10 × 45 (both sides)

Number of cells

42

32

88

12

36

Type of cells

Non-flexible

Flexible

Flexible

Flexible

Flexible

Voltage produced (V)

25.2

16

44

6

18

2.4.2

Maximum Power Point Tracking System

A maximum power point tracking system acts as an electronic DC to DC converter and particularly optimizes the suitable matching between the solar array and the battery bank. They dynamically track the maximum power point to extract maximum power output from the solar panels. Charge controllers can work with various range of power and voltage under solar photo voltaic panels [19, 20]. The specifications of the maximum power point tracking system used in the vehicle are given in Table 6 Battery charging algorithm primarily comprises of three steps, namely Boost, Absorption, and Trickle charge which gets implemented so as to charge the battery with precise current and voltage which ultimately yields fast battery charging and ensures longer shelf life of a battery.

2.5 Hardware Implementation We have divided our hardware circuit configuration into two broad categories, namely high tractive system and low tractive system. The high tractive system consists of motor, motor controllers, battery management system, and solar panels as well as a solar charger. The high tractive system is being driven by a 48 V battery. The low tractive system consists of a horn, brake light which are being driven a separate 12 V power supply. The components of the dashboard being the ignition switch, kill switch, battery level indicator, and speedometer. Solar panel extracts the solar energy and solar charger ensures the optimum transfer of solar energy which in turn energies the battery. Battery management system ensures the effective working of battery for best performance. Three-kill switches along with an MCB are mounted at suitable positions to ensure the safety of the individual. Current from the battery drives the motor whereas the movement of motors is maintained via the motor controller. A combination of hall sensors allows the movement of the rotor of a BLDC motor via hall effect. Reversing switch in hand throttle ensures clockwise or anticlockwise movements of rotor depending on the connection. Clockwise and anticlockwise movements of the rotor is associated with a forward and backward movement of the electric vehicle.

Design and Real-Time Implementation of Economical …

391

Table 6 Maximum power point tracking system Specifications Parameters

Specification

Charge controller type

Microprocessor-controlled high-efficiency PWM design

Battery voltage

72–240 V (factory set)

Maximum solar panel voltage

400 V

Charging method

3 stage battery charging—boost/absorption/trickle mode

Product efficiency (%)

Up to 98%

Max charging current (at 25 °C)

20 A

Parameters displayed on LCD

PV power, cumulative PV KWh, battery voltage, battery current, battery charging mode, night mode, time since dusk, ambient/battery temperature

Programmable parameters (using 3 buttons on Battery end-of-charge voltage (range: the front panel) 13.6–14.5 V) Battery trickle charge voltage (range: 13.2–14 V) Battery equalizing charge voltage (range: 14–14.8 V) Maximum battery charging current (range: 5–20 A) Power saver mode (optional accessory required) Optional accessories

Battery temperature sensor with 2 m cable Power saver box: for inverters (mains supply to the inverter is cut-off when sufficient solar energy is available. This ensures saving in electricity bills)

An overall connection of all these components in a circuit depicted in below figure ensures optimum performance and safety of the individual to the uttermost extent. Hardware depiction of the high tractive system has been illustrated in Fig. 5 (Table 7).

3 Result and Observations We were successfully able to design and implement a low cost less weight solarelectric hybrid car. By research and analysis, we were able to get together the best components and material required for the implementation of this project as discussed in this paper. The Solar/Electric Powered Hybrid Vehicle (SEPHV) is the solution to most of today’s problems pertaining to the environment. Figure 6 shows the CO2 emission comparison between different types of transportation [21, 22]. This vehicle

392

N. N. Kassim et al.

Fig. 5 High tractive circuit Table 7 Weight specifications Sl. no

Parts

1

Altered vehicle (without engine, fuel tank, and unwanted mechanical parts)

2

Battery bank

3

Rider (2 persons)

4

Motor

5

Solar panels

6

Miscellaneous

Fig. 6 Carbon dioxide emission comparison

Specifications (kg) 33 ± 2 35 ± 2 150 ± 10 15 ± 1.5 7±1 50 ± 10

Design and Real-Time Implementation of Economical …

393

Table 8 Comparison between SEPHV, electric vehicle, and conventional vehicle Specification

Vehicle SEPHV

Reva I (electric vehicle)

Tata Nano XE (conventional vehicle)

Weight

220 kg

400 kg

695 kg

Top speed

44

km/h

80 km/h

110 km/h

Mileage

101 km/charge

120 km/charge

22–24 km/l

Drive

BLDC motor

Induction motor

624 cc multi-point fuel injection petrol engine

Acceleration

0.6 m/s2

1.58 m/s2

2.30 m/s2

Pollution

Least

Less

Most

Charging time

6.5 h

4.5 h



Cost/km

Rs 0.49/km

Rs 0.64/km

Rs 4/km

can be charged from both solar and electric power. At the time of sparse sunlight, it can power itself through the current from the battery and run at a decent speed whereas in time of dense sunlight, it can utilize the solar energy and through effective battery charging maintain its speed. Thus, in this way, an effective mode of transportation has been developed which ensures the safety of the passenger and also provides a vehicle of greater efficiency at a relatively lower cost which suits their budget. The relatively advanced features of the solar car ensure its long run in India and subcontinent regions where solar energy is abundant in nature [12, 14]. Table 8 shows a comparison between the discussed SEPHV, and on-road electric vehicle Reva-I, and conventional engine vehicle.

4 Conclusion The result of the comparison of the three vehicles in table no. 8 based on various parameters is that the SEPHV is the most efficient and the lightest passenger vehicle among the three. Also, it has the least cost per kilometer among the three. It has the least top speed but also causes minimum pollution which makes a SEPHV ideal for metropolitan cities. Moreover, it has a minimum manufacturing cost as compared to the other two vehicles. It saves a lot of money in the long run and apart from that, it is constantly run by a source of clean energy which does not pollute the environment thus provides benefits on all fronts. The solar panels used in the car are durable and long-lasting which allow them to be used effectively for a longer duration of time [19]. The lightweight and peak performance of lithium-ion battery adds to the list of benefits. Thus, an overall solar hybrid electric car is suitable by all means to be used by the general public for automation on a day-to-day basis. The solar panels are not so efficient and thus less energy transfer takes place at times. Apart from that, there

394

N. N. Kassim et al.

are places where solar irradiance is low and solar car may not be of great use at these places. Moreover, solar energy cannot directly power the car, instead, it energizes the battery bank. Since the pros are generally much higher than cons, we can thereby conclude that solar hybrid electric car is a tenable option in the field of automation.

References 1. Manivel MV, Krishnamoorty AAP, Sudhan RN, Sidhaarth K, Sasiraju AS, Ramesh S (2017) Effect of preheating temperatures on impact properties of Chromoly alloy steel 4130 weld using gas metal arc welding. Int J Civil Eng Technol 8:319–327 2. Ramesh S, Sasiraaju AS, Sidhaarth K, Sudhan Rajkumar N, Manivel Muralidaran V (2014) Experimental investigation and hardness analysis of chromoly steel multipass welds using GMAW. World Acad Sci Eng Technol Int J Mater Metallurg Eng 8(12) 3. Baldursson S (2005) BLDC motor modelling and control—A Matlab®/Simulink® Implementation (2005) 4. Zhu ZQ, Howe D (2007) Electrical machines and drives for electric, hybrid, and fuel cell vehicles. Proc IEEE 95:746–765 5. Krishnan R, Electric motor drives modeling, analysis, and control 6. Paranjothi G, Manikandan R (2014) Photovoltaic based brushless DC motor closed loop drive for electric vehicle. Int J Emerg Trends Electric Electr (IJETEE). 10(1). ISSN: 2320–9569 7. Gerssen-Gondelach SJ, Faaij APC (2012) Performance of batteries for electric vehicles on short and longer-term. J Power Sourc 212:111–129 8. Catenacci M, Verdolini E, Bosetti V, Fiorese G (2013) Going electric: expert survey on the future of battery technologies for electric vehicles. Energy Policy 61:403–413 9. Akhter R, Hoque A (2006) Analysis of a PWM boost inverter for solar home application. . Proc World Acad Sci Eng Technol 17:212–216 10. Shaheen S (2004) California’s zero-emission vehicle mandate. Institute of Transportation Studies, UCD-ITS-RP-04-14 11. Matsumoto S (2005) Advancement of hybrid vehicle technology. In: Proceedings of IEEE European conference on power electronics and applications, Dresden, 11–14 Sept 2005, pp 1–7 12. Lalouni S, Rekioua D, Rekioua T, Matagne E (2009) Fuzzy logic control of standalone photovoltaic system with battery storage. J Power Syst 193:899–907 13. Larminie J (2003) Electric vehicle technology explained. Oxford Brookes University, Oxford/John Lowry Acenti Designs Ltd., Wiley 14. Huang KD, Tzeng S-C, Ma W-P, Wu M-F (2005) Intelligent solar-powered automobileventilation system. Appl Energy 80:141–154 15. Matsumoto S (2005) Advancement of hybrid vehicle technology. In: Proceedings of IEEE European conference on power electronics and applications, pp 1–7 16. Underland NMTM, Robinson WP (2002) Power electronics converters application and design, 3rd edn. Wiley, Hoboken 17. González-Longatt FM (2005) Model of photovoltaic module in Matlab, II CIBELEC 18. Daniels MW, Kumar PR (2005) The optimal use of the solar power automobile. Control Syst Mag IEEE 19(3) 19. Li X, Lopes LAC, Williamson SS (2009) On the suitability of plug-in hybrid electric vehicle (PHEV) charging infrastructures based on wind and solar energy. 2009 IEEE power and energy Society general meeting 20. Emadi A, Rajashekara K, Williamson SS, Lukic SM (2005) Topological overview of hybrid electric and fuel cell vehicular power system architectures and configurations. IEEE Trans Veh Technol 54(3):763–770

Design and Real-Time Implementation of Economical …

395

21. Granovskii M, Dincer I, Rosen MA (2006) Economic and environmental comparison of conventional, hybrid, electric and hydrogen fuel cell vehicles. J Power Sourc 159(2):1186–1193. https://doi.org/10.1016/j.jpowsour.2005.11.086 22. Greenhouse Gases, Regulated Emissions, and Energy use in Transportation (GREET 1.6) model. Transportation Technology Research and Development Center, Argonne National Laboratory 23. Aharon I, Kuperman A (2011) Topological overview of powertrains for battery-powered vehicles with range extenders. IEEE Trans Power Electron 26(3):868–876. https://doi.org/10.1109/ tpel.2011.2107

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids G. Sree Lakshmi, Rubanenko Olena, G. Divya, and I. Hunko

Abstract Electric vehicles (EVs), the prominent distributed energy sources (DER) can act as important ancillary services which can be used to balance the gird demand and supply during peak time. EVs provide economic benefits with required energy demand. EVs can reduce the emission of toxic gases into the environment and demand for oil resources. To bring the new technologies and intelligent controls into existing, the integration of EVs with renewable energy sources (RES) and grid plays an important role. The market is bringing new EVs and many other transportation vehicles (TV) into picture with larger battery storage capacity. This new improvements are making capable of machine learning and optimization techniques into reality. EVs with bi-directional battery storage capability can be connected to the grid and for many other specific applications, which are called as gridable EVs (GEVs). In this paper, challenges and importance of grid-connected EVs, such as vehicle-to-vehicle (V2V), vehicle-to-home (V2H), vehicle-to-load (V2L), and vehicle-to-grid (V2G) technologies are discussed. Keywords Electric vehicles (EV) · Distributed energy resource (DER) · Renewable energy sources (RES) gridable electric vehicles (GEV) · Vehicle-to-vehicle (V2V) · Vehicle-to-home (V2H) · Vehicle-to-load (V2L) · Vehicle-to-grid (V2G)

1 Introduction In today’s scenario, the two important issues that are running in every discussion are ‘Energy’ and ‘Environmental Pollution.’ The global system prominent need is a G. Sree Lakshmi (B) · R. Olena Faculty of EE, RICE, UWB, Pilsen, Czech Republic e-mail: [email protected] G. Divya Electrical and Electronics Engineering, CVR College of Engineering, Hyderabad, India I. Hunko PE and Electro Mechanics, Vinnytsia NTU, Vinnytsia, Ukraine © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_30

397

398

G. Sree Lakshmi et al.

clean, efficient, intelligent, and environmental friendly transportation system (TS). Electric vehicle technology has taken an accelerated path and gained popularity from industries, government, and customers [1, 2]. EVs are gaining importance in the world by reducing the dependency of fossil fuels and making it economical by reducing the cost of the transportation. EVs are becoming the solution for improving the quality of air by reducing the gas emission into the environment. The concept of E-mobility is not a new idea, but still making a fast move in making the world sustainable. It can able to meet the challenges of the energy security. E-mobility technology is an integration of vehicle body, battery energy storage, electric propulsion, and energy management together [3–6]. In past, EVs are focused on individual components such as motor, drive used, batteries, fuel cells, and renewable energy sources, but now, the research is taking place on the how efficiently that EVs can be used for various applications in exchange of energy with the power grid. EVs when used for exchange of energy with the power grids are called as gridable EVs or grid-connected EVs (GEVs). They can able to draw energy from the grid and can also supply energy to the grid during peak demand. The transformation of energy from grid to EVs and EVs to grid can be possible by using an important component called bi-directional charger. It consists of DC link capacitor which can supply reactive power to the grid [7–10]. Conventional power plants which uses fossil fuel has very low efficiency (30%) because of various constraints but however modern power plants using renewable energy sources can give higher efficiency (70%) [11]. The only problem with the renewable sources is their intermittent nature due to environmental conditions. This nature adversely affects the frequency, grid voltage, and the reactive power of the power system. There is a need for continuous compensation or regulation of power grid. This compensation will become crucial when utility home is integrated with the small-scale renewable energy generation [12, 13]. This can be overcome by a single GEV connected to the utility home. A group of GEVs can also be connected together to support the community-grid operation. Based on the charging and discharging capabilities of the GEVs and efficient-energy requirement of the power grid, various concepts came into existence such as vehicle-to-home (V2H), vehicle-to-grid (V2G), and vehicle-to-vehicle (V2V). Basically V2H, V2G, and V2V not only enable GEVs to form a transportation tool but also they can act as controllable loads and distributed energy sources (DES) for the power grid. For all this energy transformations, bidirectional charger is needed which can provide additional reactive power to the system from the capacitive devices to support the inductive devices demanding high reactive power during operation [12, 13]. This paper investigates about the challenges and opportunities of new GEVs technologies such as V2H, V2L, V2G, and V2V connected to the power grid. The importance of renewable energy sources integrating with the electric vehicles is discussed. Ecological and economical benefits of EVs integrated with RES and smart grid. The buffer charging criteria for the renewable energy sources integrating with the electric vehicles are discussed [14, 15]. The challenges of renewable energy sources integrating with gridable electric vehicles are discussed. Overall, there is a great need of integrating electric vehicles with

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

399

renewable energy sources for decreasing the pollution in the environment and making the world sustainable.

2 Renewable Energy Sources The increasing awareness of environmental crisis and fossil fuels depletion drives for other promising alternatives, which brings RES for environmental friendly and sustainability.

2.1 Solar Power Energy Energy can be generated from solar plants by using photovoltaic solar panels and concentrating solar power plants. Concentrating solar power plants utilizes mirrors to focus sunlight to create high temperature to drive traditional steam turbine to generate electricity. The PV solar power plants generate electricity by using solidsate semiconductor which converts sunlight energy into electrical energy. The important element in the PV solar panel is the solar cell which runs efficiently at low ambient temperature. The power efficiency of these solar cells goes on decreases as the temperature increases. The decrease in efficiency is equal to 0.2–0.5%, for the increase in temperature of 1 °C. To avoid this decrease in efficiency of solar cell with the increase in temperature, the manufactures adopt heat insulation technique. The output of the solar cell depends upon the intensity of solar irradiation and the area of aggregated solar cells and given as: PSolar (s) = ηsolar

(1)

The electricity output of the solar PV panels is given as PPV (t) = Aβμ(t)

(2)

where ‘A’ is the area of the PV panel, β is the efficiency of the PV panel, μ(t) is the insolation. The most important issues are the modeling of solar energy. Khatib survey has given model for linear, nonlinear, and artificial intelligent model for solar energy prediction. The linear global solar energy model is given as [2]: ET =a+b E extra

(3)

where E T represents the global solar energy, E extra represents the extra terrestrial solar energy, S represents the day length S o represents the number of shining hours,

400

G. Sree Lakshmi et al.

‘a’ and ‘b’ are coefficients of the model. The commonly used diffused solar energy linear model is given as [2]: Ed =a+b ET

(4)

where E d represents the diffused solar energy, E T represents the global solar energy and K T represents the clearness index, ‘a’ and ‘b’ are the coefficients of the model [2].

2.2 Wind Power Energy Wind energy sources are playing a vital role for supplying renewable energy. For efficient generation of wind energy, choosing the location of wind farms is very important. The quality and quantity of the wind energy should be assessed which is directly related to the wind speed. It depends upon the height of the wind turbine. The expression for selection of the height of the wind turbine is given by Peterson and Hennessey [18]: 

v vr



 =

h hr

a (5)

Here, ‘ν’ is the wind speed, ‘h’ is the height, ‘vr ’ is the known wind speed and ‘hr ’ is the reference height, and ‘a’ is the coefficient which varies from 0.1 to 0.4 [19]. The amount of energy generated by wind sources also depends upon the wind speed and air density. P=

1 αρ AV 3 2

(6)

where ‘α’ refers to the Albert Betz constant and ‘ρ’ represents the air density in kg/m3 . Additionally, the parameters of wind turbine and intermittence of the seasons also influence the electricity generated by wind energy sources [2].

3 Electric Vehicles Integration with Renewable Energy Sources The electric vehicle has three important segments, such as storage unit, control unit, and impetus unit. Storage unit is the one which stores the energy and where the actual energy conversion and storage process takes place inside the battery. From this storage point, the power goes to the control unit. The controller is connected

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

401

Fig. 1 Basic components of electric vehicle

to the electric engine. The major functions of the controller unit are to regulate the power, choosing the amount of required power, converting the power from DC to AC. Then, the AC control drives the electric engine which changes the electrical power to mechanical power. This forms the impetus unit of the EV. Every EV has two engines, one for power and other for energy. The basic components of the electric vehicle are shown in Fig. 1. The main purpose of electrifying transportation is to reduce oil dependence and to reduce vehicle emissions. There are two main paths to store electricity for use in electric vehicle. The first one is by shifting oil-derived fuels to electricity-derived fuels, either liquid or gaseous, with hydrogen as most important component. Second one is to store electricity on board the vehicle using batteries. An electric vehicle is the one which has at least one electric motor in a vehicle to drive the power train. Electric vehicles are classified as below and the vehicle electrification is shown in Fig. 2. 1. 2. 3. 4. 5.

Battery electric vehicle (BEVs) Hybrid electric vehicle (HEV) Plug--n hybrid electric vehicle (PHEV) Range-extended electric vehicle (REEV) Fuel cell hybrid electric vehicle (FCHEV).

BEV mainly consists of three important parts, namely electric motor, controller, and rechargeable battery. In this case, electric motor utilizes rechargeable battery as the energy source for propulsion. A two quadrant controller is used to supply power to electric motor which drives the vehicle to move forward and backward. A four-quadrant controller is also used for regenerative braking. BEV is fully operated

402

G. Sree Lakshmi et al.

Fig. 2 Vehicle electrification

by electrical energy. The combination of battery and the re-charging features by electric outlet makes BEV a zero emission vehicle. Hybrid electric vehicle (HEV) is a combination of battery and traditional combustion engine. In this case, a battery is used to provide power for the electric motor below speed of 40 miles per hour with zero emissions and the combustion engine will drive at higher speeds. Plug-in hybrid electric vehicles (PHEV) is same as HEV with additional feature of re-charging by plugging into an electric outlet. They can have good fuel efficiency and environmental friendly characteristics. The energy diversification of EVs is shown in Fig. 3. Extended REEV is a combination of BEV and PHEV with more improved fuel efficiency and reduced CO2 emissions. Fuel cells HEV works on the fuel cell technology for charging the batteries. Transportation of electrification (TE) provides the ancillary services to the smart grid (SG), such as voltage regulation, frequency, and peak shifting. The requirement of energy storage systems arises with the large penetration of RES owing to their intermittence nature. Transportation of electrification becomes most suitable dynamic energy storage. Transportation of electrification possesses many challenges to the grid, such as controlling, planning, and operation. Advanced control schemes and communication infrastructure play a vital role for complete establishment of transportation of electrification. Renewable energy-powered EVs form the essential catalyst of our power infrastructure. The conventional energy source is the prime objective of the current scenario; renewable energy integration is the feasible solution. The charging stations and the parking lots are located in open geographical places which makes the usage of abundant free energy. The maximum amount of energy can be stored in the EVs from the renewable energy sources during the excessive generation of the sources based on the charging stations and parking slots with the minimum area consumption. The stored energy in the EVs can be utilized during peak hours for various applications.

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

403

Fig. 3 Energy diversification of EVs

4 Gridable Electric Vehicle 4.1 Types of Gridable EVs The different types of GEV technologies are V2H, V2V, V2L, and V2G. (i) Vehicle-to-home (V2H) technology: In this case, gridable electric vehicle (GEV) is used as home generator during the periods of failure of electrical services. This type of GEV is directly connected to the home for charging/discharging by the onboard or offboard bi-directional charge. It can able to draw energy from home or it can deliver energy to home based on the control algorithm. (ii) Vehicle-to-vehicle (V2V) technology: In this case, GEV is used to transfer electrical energy between another PEV in case of emergency. Bi-directional chargers are used through the local grid and then distribute the energy between the GEVs by a controller called aggregator. Aggregator plays a very important role in connecting and interacting GEVs among themselves and with the grid for the energy. (iii) Vehicle-to-load (V2L) technology: In this case, GEV is used to transfer power to a remote site or load which has no electrical connections, like construction building sites or big camps. The main advantage of this technology is availability of power

404

G. Sree Lakshmi et al.

in remote areas where there is no possibility of electricity generation. This V2L technology can help in heating, flooding, cooking process, and emergency lighting in the remote areas during the emergency tragic natural catastrophe periods. (iv) Vehicle-to-grid (V2G) technology: In this case of GEV, electric utility purchases energy from the customers owning GEV during the peak hours and use the GEV battery capacity for providing ancillary services balancing the energy production and consumption. An intelligent and collaborative system is needed to make the possibility of battery charging and discharging the energy to the grid when needed. Every GEV needs two fundamental elements such as connection for the bi-directional flow of electrical energy and a logical connection for control and communication from both sides. This connection can be set through broadcast radio signal through cell phone network from an Internet connection. Potential benefits of GEVs technologies are proven to be multiple but depend on the application. The main benefit of GEVs is the readily available to participate in the power market in providing services to the grid. They can participate for balancing the grid services, such as voltage stability and frequency regulation. They can accrue revenue and return on investment of the vehicles to the owner. The increasing penetration of renewable such as wind and solar, GEVs can serve as a potential storage, allowing better renewable integration into the grid (V2G and V2H applications). Renewable energy sources play a key role in development of GEVs market because EVs can take increase in load from renewable energy sources during excessive generation and helps in balancing generation with demand and stabilizing the intermittency production. This makes the cleaner and cheaper electricity. The GEVs power flow can be possible in two modes, such as unidirectional and bi-directional. The unidirectional GEVs control the EVs battery charging time in one direction of power flow. This mode of operation is simple, easy, and inexpensive. It generates trade policies between power utilities and the owner of the EV. The owner has to embolden to the policies of the energy needs. It provides a profit minimization and maximization for the utility and consumers. The main problem with unidirectional mode of operation of GEVs is trapping of power during peak loads, reactive power support, frequency, and voltage regulations balancing. The bi-directional GEVs have power flow in both the directions achieving maximum benefits. They contain bi-directional converters AC/DC and DC/DC converters. Choppers are also used to control bi-directional power flow. The bidirectional GEVs provide greater flexibility and opportunities to improve the power system stability. The main advantages of bi-directional GEVs are active, reactive power support, power factor control, frequency and voltage stability, support for the integration with the RES.

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

405

4.2 Ecological and Economical Benefits In general, maximum quantity of power is contributed to the grid by the coal combustion but when it is generated by RES, one can reduce the emissions. Charging of EVs using solar power can reduce 0.6 ton of CO2 per car per year which reduces the carbon emissions by 55% when compared to the charging of EV at home during night. By using optimal control strategies for EVs and PV integration, another 0.36 ton CO2 can be reduced [2]. When EVs are integrated with RES reduces carbon emission into the environment. The integration of EVs with RES also provides economic benefit for charging providers, EV owners, and power grid enterprises. For charging providers, the integration of EVs and RESs is influenced by operation cost. The operation cost of power system can be modeled based on unit commitment problem which is defined as [2], OC =

T N   [FCit (Pit ) + MCit (Pit )] + STit + S Dit

(7)

i=1 t=1

Here, OC is the operational cost, FCit (Pit ) is the fuel cost, MC it (Pit ) is the maintenance cost, STit is the start-up cost and SDit is the shut-down cost. By considering the intermittence of solar and wind power generation, and the randomness of EV load, the operational cost can be determined. The operation cost can be minimized by proper integration of grid with RES and EVs. EV owners also can save lot of amount by following the guidelines of the energy sector, when their EVs are integrated with grid and RES.

4.3 Efficient Operation of EV Charging Stations A lot of electric cars have the limited mileage, near 100–300 km. For the widespread introduction of electric vehicles, it is necessary to build along the highways of the charging stations. Nowadays, charging stations for electric vehicles powered by existing power grids are concentrated mainly in large cities. Obviously, the use of electricity from traditional energy sources (primarily thermal power plants) to charge electric vehicles will not reduce harmful emissions. The unreliable supply of electricity to end consumers (charging stations) is confirmed by the high grid defect ratio of 13% in Ukraine, as well as the operation of transmission lines, which are modernized slower than gained. Frequent emergencies with power grids have become common place. And when in large cities, this problem is solved relatively quickly, small towns and villages can be held hostage for a long time. Voltage fluctuations in the electrical network in small cities and towns are shown in Fig. 4 [14] and average frequency deviation in the electrical network is shown in Fig. 5 [15].

406

G. Sree Lakshmi et al. U, V 260

Unaccepteptable Voltage Deviation

250

Maximum Allowable Voltage Deviation

240 230

Allowable Voltage Deviation

220 210 200 190

Unaccepteptable Voltage Deviation

180 170 160 00:00

04:00

08:00

12:00

20:00

16:00

24:00

T, hours

Fig. 4 Voltage fluctuations in the electrical network in small cities and towns

Average frequency deviation by 08.11.2019 30

Duration per day,%

25

20

15

10

5

0 -0.06

-0.05

-0.04

-0.03

-0.02

-0.01

+0.01

+0.02 +0.03

+0.04

+0.05

+0.06

Rejection of 50.00 Hz

Fig. 5 Average frequency deviation in the electrical network

Only the use of renewable energy sources (RES) to charge electric vehicles can replace traditional energy sources in the country’s balance sheet and give the desired effect. In addition, the use of RES will provide the necessary dispersion of generating capacity and will make the charging stations independent of the grid. Given the irregularity of RES generation, a comprehensive approach to their use is needed. The new way into the perspective of using renewable energy sources is to research of charging station for electric vehicles. The primary generator of electricity for charging plants for electric vehicles, use power plants that convert energy from renewable energy into electrical energy. The realization of charging stations using solar radiation and wind need to decide a lot of problems related to the establishment of specifications of charging stations of this type, a mathematical description of the processes of conversion and storage of energy.

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

407

The uneven nature of the electricity generation of wind turbines and photovoltaic power plants, both daily and monthly and seasonally, on the one side, and the random nature of electricity consumption by electric vehicles on the other side is a big problem. Efficient operation of the charging station from the wind turbines and photovoltaic power plants in terms of balancing the generation and consumption power is possible, provided to its composition by buffer battery, how it is shown on Figs. 6 and 7 .[16]. Taking into account random power generation by renewable sources, which is the place in the daylight hours, better to use charging station of a

Fig. 6 Wind-based charging station

Fig. 7 Solar-based charging station

408

G. Sree Lakshmi et al. Solar Energy [kWh]

Wind Energy [kWh]

Giraffe Energy [kWh]

1200

900

800

600

400

200

0 Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Fig. 8 Generation of electric energy by a complex (wind and sun) charger station

combination wind-sun plant and a buffer battery. Generation of electric energy by a complex (wind and sun) charger is shown in Fig. 8. [16]. The battery power of an electric vehicle can be shown as [17]: E CCTBEV = E RWTAB + E STABEVVM

(8)

where: E RWTAB

E STABEVVM

is the residual energy of the traction AB of the electric vehicle remaining in it at the time of the beginning of the process of charging on the WEU from the combine solar-wind plant, kWh. is the spent (discharged) energy of the traction AB of an electric vehicle, the process of vehicle movement, kWh.

During the operation of such a charging station, when the charging process of the traction AB of an electric vehicle is guaranteed by a buffer battery, which equates random power generated with random consumption, the energy of the buffer battery can be shown as:

E BufAB

E BufAB = E RbufAB + E SbufTABEV ;

(9)

E WPP + E PPP = E BufAB + E RWTAB ;

(10)

is the energy of the buffer battery that is included autonomous windsolar-driven electric vehicle charging station, kWh;

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

E RbufAB E SbufTABEV

409

residual energy remaining in the buffer battery after charging the electric vehicle, kWh; energy spent on charging the traction AB of an electric vehicle, kWh.

5 Challenges of Renewable Energy Sources Integration with Gridable EVs The modelling of gridable EVs (GEVs) such as V2H, V2L, V2V, and V2G should be based on their objectives and constraints. The general objectives of GEVs are load variance minimization, cost minimization, cost efficiency optimization, costemission minimization, power loss minimization, load shift and peak load reduction, and reactive power compensation. Based on the different sizes of GEV batteries and capacitor, V2H, V2L, V2V, and V2G technologies have their individual objectives and constraints. The grid-integrated EV charging system has major impact on power quality of the grid. When large number of vehicles are connected to the grid for charging simultaneously, then the grid will get overloaded. This situation will distort the power quality of the grid. This can be overcomed by integrating renewable energy sources with the charging stations, which can avoid the overload condition of the grid and maintains the grid stability. The design and control mechanism play an important role when GEVs are integrating with the RES. The periodical control monitoring is required for proper flow of power from grid to vehicles and vehicle to grid using renewable sources. A bi-directional converter is properly designed to make the power flow in both the directions. A selection of battery also plays a major role when the GEVs have to integrated with RES. The battery should have a capability of good charging and discharging, such that they can exactly match with the voltage and current ratings of the grid. The battery should also have less degradation rate. The size of the batterys should also be flexible to the applications. The main important factor that should be considered for integrating GEVs with RES and grid is the ratings of the them should match. The communication between the grid, RES, and the GEVs should be monitored continuoulsy for efficient usage of their services. The V2G telematics plays an important role in transfering the data wirelessly over long distances with good reliability. By proper maintaining the communication and control strategy between GEVs, RES, and grid, an effective power flow mechanism can be generated between V2V, V2H, V2L, and V2G technologies.

6 Conclusions Electric vehicle penetration is reshaping the transportation system. Integration of EVs with renewable energy sources and the grid called gridable EVs (GEVs) is playing an important role in this modern power system world of power generation. GEVs like V2V, V2H, V2L, and V2G when connected to grid will act either like source or load

410

G. Sree Lakshmi et al.

depending upon the application and the need. GEVs when properly coordinated and communicated with the RES and grid will able to provide the ancillary services like voltage stability, frequency control, reactive power compensation, load balancing, and enhancing renewable power penetration to the grid. The task of integrating the GEVs with the grid and RES supporting customer preferences and creating benefits for utility companies is technically and economically challenging. In this paper, it has clearly discussed about the importance of integration of GEVs with RES and grid. The details of different types of EVs, renewable energy sources integration, ecological and economical cost benefits, buffer charging of EVs using solar and wind are clearly discussed. The analysis clearly states that GEVs when integrated with RES and grid are becoming the one of the best ways in balancing the power generation and utilization problem with good environmental benefits saving the people from fossil-fuel pollution.

References 1. Dyke KJ, Schofield N, Barnes M (2010) The Impact of Transport electrification on Electrical Networks. IEEE Trans Ind Electr 57:3917–3926 2. Liu L, Kong F, Liu X, Peng Y, Wang Q (2015) A review on electric vehicles interacting with renewable energy in smart grid. Renew Sustain Energy Rev 51:648–661 3. Richardson DB (2013) Electric Vehicles and the electric grid: a review of modeling approaches, impacts and renewable energy integration. Renew Sustain Energy Rev 19:247–254 4. Tan KM, Ramachandaramurthy K, Yong JY (2016) Integration of electric vehicles in smart grid: a review on vehicle to grid technologies and optimization techniques. Renew Sustain Energy Rev 53: 720–732 5. Ehsani M, Falahi M, Lotfifard S (2012) Vehicle to grid services: potential and applications. Energies 5:4076–4090 6. Corchero C, Sanmarti M (2018) Vehicle-to-everything: benefits and barriers. In: IEEE conference proceedings, 2018 7. Monteiro V, Goncalves H, Ferreira JC, Afonso JL (2012) Batteries charging systems for electric and plug-in hybrid electric vehicles. Intech, pp 149–168 8. Dai J, Dong M, Ye R, Mo, Yang W (2016) A review on electric vehicles and renewable energy synergies in smart grid. IEEE conference proceedings, CICED 9. Esther S, Singh SK, Goswami AK, Sinha N (2018) Recent challenges in vehicle to grid integrated renewable energy system: a review. In: IEEE conference proceedings ICICCS-2018, pp 427–434 10. Rizvi SAA, Xin A, Masood A, Iqbal S, Jan MU, Rehman H (2018) Electric vehicles and their impacts on integration in to power grid: a review. In: IEEE conference proceedings, 2018 11. Liu C, Chau T, Wu D, Gao S (2013) Opportunities and challenges of vehicle-to-home, vehicleto-vehicle, and vehicle-to-grid technologies. Proc IEEE 101(11):2409–2427 12. Lopes JAP, Soares FJ, Almeida PMR (2011) Integration of electric vehicles in the electric power systems. Proc IEEE 99(1):168–183 13. Saber AY, Venayagamoorthy GK (2011) Plug-in vehicles and renewable energy sources for cost and emission reductions. IEEE Trans Ind Electr 58:1229–1238 14. Budko VI (2019) Utilizing solar and wind power to charge electric vehicles. Manuscript rights. Doctoral thesis in specialty 05.14.08—Renewable energy transformation. National Technical University of Ukraine—Igor Sikorsky Kyiv Polytechnic Institute, Institute for Renewable Energy, NAS of Ukraine, Kyiv

Electric Vehicles Integration with Renewable Energy Sources and Smart Grids

411

15. Integrated Power System of Ukraine Operation. Available: https://ua.energy/activity/dispatchinformation/ues-operation/ 16. https://www.innoventum.se/giraffe-2-0/ 17. Pavlov VB, Budko VI (2017) Charging of electric vehicles from renewable energy sources. Tech Electrodyn 6:32–35 18. Peterson EW, Hennessey JP (1978) On the use of power laws for estimates of wind power potential. J Appl Meteorol 19. Masters MC, Fernandez E (2013) Renewable and efficient electric power systems. Wiley, Hoboken

Allied Technologies

FOPD Controller Using Bee Colony Optimized Reduced Order FOFOPDT Model of Three Interacting Tank Process U. Sabura Banu, Abdul Wahid Nasir, and I. Mohamed Shiek Mothi

Abstract In this paper, fractional-order proportional-derivative (FOPD) controller is designed based on three frequency domain design criteria for the bee colony approximated fractional-order first-order plus dead time system (FOFOPDT). Three tanks connected interactively in series are considered for the study. As the process taken is a nonlinear process, local linear FOFOPDT model is developed for each region separately minimizing root mean square error using bee colony optimization algorithm. Time and frequency response of the original process is compared with the approximated model for validating the model. Since there are three controller parameters for both FOPD controller, three design criteria are fulfilled. Here proportional gain, derivative gain, and order of the differentiator are the three controller parameters of the FOPD controller. Simulation results under servo, regulatory, and parameter variations are reported. Keywords Bee colony optimization · FOFOPDT system · Fractional-order proportional-derivative controller · Robustness to parameter variation

1 Introduction In many process industries such as petrochemical, pharmaceuticals, food and beverage industries, the process fluid is stored in cylindrical tanks. The fluids in the tanks are processed and transported to the next stage of the process. The tanks are connected either in interactive or non-interactive manner. When more than two tanks are connected, they become a higher-order process with mild nonlinearity, whose effective modeling and control will boost the economy of the industry. Systems are U. Sabura Banu (B) Department of Electronics and Instrumentation Engineering, B.S. Abdur Rahman Crescent Institute of Science and Technology, Vandalur, Tamilnadu, India e-mail: [email protected] Abdul Wahid Nasir NIT, Jamshedpur, India I. Mohamed Shiek Mothi Eswari Engineering College, Chennai, India © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_31

415

416

U. Sabura Banu et al.

normally modeled and controlled in integer-order form. In recent past, fractional calculus is gaining popularity in almost all fields of science and technology [1]. Using fractional calculus, the order of the integrator or differentiator is fractional, i.e., real instead of integer. Using fractional calculus in modeling, one more parameter is added in addition to the three parameters in FOPDT system. In PID controller with the inclusion of fractional calculus, two more terms are introduced thereby increasing the flexibility of the modeling and controller design. General representation of fractional-order PID takes the form PIλ Dμ which is an extension of the integerorder PID where μ and λ are the order of the differentiator and integrator, respectively. Simplification of PIλ Dμ controller is fractional-order PI[PIλ ] controller [2] and fractional-order PD [PDμ ] controller [3, 4]. Straight forward approaches are used to mathematically obtain the three parameters in both FOPI and FOPD controller. Either time domain or frequency domain specifications such as rise time, settling time, peak overshoot, phase margin, gain crossover frequency, robustness to the gain variation [5], or graphical method which involve plotting the stabilizing boundary curves in the parameter space of the controller [6] can be used. FOPI controller design using fractional MIGO-based tuning rule for Pade’s approximated first-order plus dead time (FOPDT) is reported [7]. For integer-order systems, FOPI and FOPD controllers are tuned considering specified phase margin (øm ), gain crossover frequency (ωgc ), and robustness criteria that are discussed [8, 9]. Considering minimal IAE and sensitivity constant, integer- and fractional-order PID controllers are designed for FOPDT system [10]. Two schemes of FOPI controllers for a class of fractionalorder systems [2] are discussed. Fractional-order delayed system with conventional integer-order PID controller is attempted [11]. Lead compensator design for a timedelay system to reach gain and phase margin is designed [12]. A graphical tuning method for fractional-order PID (PIλ Dμ ) controllers is proposed [13] based on the sensitivity function constraint of the closed loop, which provides the information on robustness to plant uncertainties. Optimization techniques are also investigated for the tuning of the controller parameters. Advanced multiobjective optimization using pareto optimum design for integer-order and fractional-order PID controllers for both integer-order and fractional-order plants with parametric uncertainties is discussed [14]. Relation between various fractional-order (FO) PID controllers optimally tuned based on the time and frequency domain tuning is proposed for higher-order processes [15]. Suboptimal PID controller [16] tuning methodology is used along with fractional-order pole-zero placement approach. An improved evolutionary non-dominated sorting genetic algorithm (NSGA-II), augmented with a chaotic Henon map, used for the multiobjective optimization-based design procedure is proposed [17]. Evolutionary optimization technique for model order reduction in frequency and time domain is dealt with for tuning PID and PIλ Dμ controllers [18]. Swarm intelligence techniques like bee colony optimization, ant colony optimization, bacterial foraging, etc., are used in almost all the domain for getting the global optimal solution in the recent past [19, 20]. Seeley [21] proposed a behavioral model of self-organization for a colony of honey bees. Implementation of fractionalorder controllers for multivariable process is presented in recent research [22–25]. Section 2 discusses about three interacting tank process, mathematical modeling,

FOPD Controller Using Bee Colony Optimized Reduced …

417

state-space formulation and computation of transfer function, and bee colony optimization for the formulation of FOFOPDT system. Section 3 discusses about the gain and phase computation of FOFOPDT system. Section 4 elaborates the design of FOPD controller considering frequency domain specification for FOFOPDT system. Section 5 discusses the results and conclusion.

2 Three Interacting Tank Process In most of the industries, interacting tanks connected in series are available. In this research, three similar cylindrical tanks with working fluid as water and with equal cross-sectional area (A) are considered. Connection between the tanks is through cylindrical pipes of the same cross-sectional area (α). Input fluid flow (F in1 ) is given to tank 1 from the sump by motor pump through a control valve. Through the interconnecting pipes, the liquid from tank 1 enters tank 2 and from tank 2 to tank 3. Differential pressure transmitters are used for measurement of level in each tank. The flow of liquid through the interconnecting tank is fixed by a manual value in predetermined condition. Additionally, an output pipe with manual valve is placed at the bottom of each tank. Height of the third tank is the desired output which needs to be controlled by manipulating the inflow F in1 of tank 1. Figure 1 shows the schematic diagram with the process tanks, actuators, and the level transmitters. Process parameters A1 = A2 = A3 = 615.75 cm2 α1 = α2 = α3 = 5.0671 cm2 β12 = 0.9, β23 = 0.8, β3 = 0.3 K 1 = K 2 = 75 cm3 /vs. Fig. 1 Three tank interacting system

418

U. Sabura Banu et al.

2.1 First Principle Model of Three Interacting Tank System Three interacting tank process discussed above is mathematically modeling using the first principle equations. Applying mass balance equation and Bernoulli’s principle, the dynamics of the three interacting tank is modeled as nonlinear ordinary differential equations (ODE). Mathematical model of the three interacting tank process is as follows F1 =

dh 1 (t) K1 β12 α12  2g(h 1 (t) − h 2 (t) + u1 =− dt A1 A1

(1)

F2 =

β12 α12  dh 2 (t) β23 α23  = 2g(h 1 (t) − h 2 (t) − 2g(h 2 (t) − h 3 (t) dt A2 A2

(2)

F3 =

β23 α23  dh 3 (t) β3 α3  K2 =− u2 2g(h 2 (t) − h 3 (t) − 2gh 3 (t) + dt A3 A3 A3

(3)

where hi ui Ai αi βi j Ki g

Height of tank i (cm). Manipulated variable to control valve cvi (v). Area of tank i (cm). Cross-sectional area of interacting pipe i (cm2 ). Valve opening of the interacting pipe. Pump gain (cm3 /vs). Gravity.

Nonlinear ordinary differential equation first principle models are represented as state-space models with h 1 , h 2 and h 3 being the states. But due to the presence of nonlinear terms in the equations, the equations must first be linearized. The statespace realization of the system is much more acceptable than any other realization, because of the high interaction found between the subsystems (i.e., tanks). Piecewise linearization of the system is done about different operating points. The determination of different operating points is obtained from the open-loop response of the system. The input--output characteristics are given in Table1 and Fig. 2. The I/O characteristics plot is split into three regions about whose operating points the system is linearized. The linearized state-space model can be represented as X˙  = AX  + BU

(4)

Y  = C X  + DU 

(5)

X  = X − X s , U  = U − Us , Y  = Y − Ys

(6)

FOPD Controller Using Bee Colony Optimized Reduced … Table 1 I/O characteristics

419

u 1 (V)

h 1 (cm)

h 2 (cm)

h 3 (cm)

0.5

1.33

1.285

1.24

1

3.1

2.955

2.75

1.5

5.67

5.37

4.955

2

9

8.49

7.55

2.5

13.1

12.27

11.18

3

18

16.55

15.49

3.5

23.54

22

19.99

4

30

28

25.25

4.5

37.34

34.5

31

5

45.4

42

37.5

5.5

54

50

44.53

6

63.7

58.8

52.40

6.5

74

68.3

60.55

7

85

78.47

70

7.5

96.91

89.1

79.4

Fig. 2 I/O characteristics

where X Y U X U Y Xs

State of the three interacting tank process [h 1 , h 2 , h 3 ]. Output of the three interacting tank process [h 1 , h 2 , h 3 ]. Input of the three interacting tank process [U1 , U2 ]. State deviation (difference between the current states and the steady-state values of the states). Control deviation (difference between the current inputs and the steady-state values of the inputs). Output deviation (difference between the current outputs and the steady-state values of the outputs). Steady-state operating point of the state around an operating region.

420

U. Sabura Banu et al.

Ys Output operating point around an operating region. Us Input’s operating point around an operating region.

2.2 State-Space Formation for Three Interacting Tank Process Operating conditions are obtained using piecewise linearization. Range and the steady-state values for each linear region are found out. Jacobian approximation is used to get the state-space model of the process around each region by linearizing the nonlinear ODEs. State-space model linearized for each region is tabulated in Table 2. Using state space to transfer function conversion, third-order transfer function is obtained for the individual state-space model and tabulated in Table 3. ⎡ ⎢ ⎢ A=⎢ ⎣

 − β12Aα1 12  β12 α12 A2

g 2(h 1 s−h 2 s)

g 2(h 1 s−h 2 s)

β12 α12 A

β23 α23 A3

0



g 2(h s−h s)

1 1 2 g − β12Aα2 12 2(h 1 s−h 2 s) +  g 2(h 2 s−h 3 s)

β23 α23 A2



g 2(h 2 s−h 3 s)



0 β23 α23 A2

⎤ 

g 2(h s−h s)

 2 3 g − β23Aα3 23 2(h 2 s−h 3 s) +

⎤ 0 ⎥ ⎢ B=⎣ 0 0 ⎦ K2 0 A3 ⎡

β3 α3 A3



g 2h 3 s

⎥ ⎥ ⎥ ⎦

(7)

K1 A1

(8)



⎤ 100 C = ⎣0 1 0⎦ 001

(9)

D=0

(10)

Table 2 State-space model for different regions 0.5–2

2–3.5

3.5–4.5

A ⎡

−0.3579 ⎢ ⎢ 0.3579 ⎣ 0 ⎡ −0.1783 ⎢ ⎢ 0.1783 ⎣ 0 ⎡ −0.1105 ⎢ ⎢ 0.1105 ⎣ 0

⎤ 0.3579

0

⎥ −0.6335 0.2755 ⎥ ⎦ 0.2755 −0.3036 ⎤ 0.1783 0 ⎥ −0.3168 0.1385 ⎥ ⎦ 0.1385 −0.1549 ⎤ 0.1105 0 ⎥ −0.1977 0.0872 ⎥ ⎦ 0.0872 −0.0981

B ⎡ 0.1218 ⎢ ⎢0 ⎣ 0 ⎡ 0.1218 ⎢ ⎢0 ⎣ 0 ⎡ 0.1218 ⎢ ⎢0 ⎣ 0

⎤ 0 0

⎥ ⎥ ⎦

0.1218 ⎤ 0 0

⎥ ⎥ ⎦

0.1218 ⎤ 0 0 0.1218

⎥ ⎥ ⎦

C ⎡ 1 ⎢ ⎢0 ⎣ 0 ⎡ 1 ⎢ ⎢0 ⎣ 0 ⎡ 1 ⎢ ⎢0 ⎣ 0

⎤ 00 1 0 0 1 0 0 1 0

⎥ 0⎥ ⎦ 1 ⎤ 0 ⎥ 0⎥ ⎦ 1 ⎤ 0 ⎥ 0⎥ ⎦ 1

D 0

0

0

FOPD Controller Using Bee Colony Optimized Reduced …

421

Table 3 Transfer function matrix for different regions Region (F in )

A ⎡

0.5–2

⎢ ⎢ ⎣

2–3.5



3.5–4.5



⎢ ⎢ ⎣

⎢ ⎢ ⎣

0.1218s 2 +0.114s+0.01418 s 3 +1.295s 2 +0.3237s+0.0028 0.04359s+0.01323 s 3 +1.295s 2 +0.3237s+0.0028 0.012 s 3 +1.295s 2 +0.3237s+0.0028 0.1218s 2 +0.0574s+0.00364 s 3 +0.65s 2 +0.0822s+0.0004 0.0217s+0.0034 s 3 +0.65s 2 +0.0822s+0.0004 0.003 s 3 +0.65s 2 +0.0822s+0.0004

0.012 s 3 +1.295s 2 +0.3237s+0.0028 0.03356s+0.012 s 3 +1.295s 2 +0.3237s+0.0028 0.1218s 2 +0.1208s+0.012 s 3 +1.295s 2 +0.3237s+0.0028 0.003 s 3 +0.65s 2 +0.0822s+0.0004 0.01687s+0.003 s 3 +0.65s 2 +0.0822s+0.0004 0.1218s 2 +0.0603s+0.003 s 3 +0.65s 2 +0.0822s+0.0004

0.1218s 2 +0.03603s+0.001436 s 3 +0.4063s 2 +0.03227s+0.0001 0.01346s+0.0013 s 3 +0.4063s 2 +0.03227s+0.0001 0.0012 s 3 +0.4063s 2 +0.03227s+0.0001

⎤ ⎥ ⎥ ⎦

⎤ ⎥ ⎥ ⎦

0.0012 s 3 +0.4063s 2 +0.03227s+0.0001 0.0106s+0.0012 s 3 +0.4063s 2 +0.03227s+0.0001 0.1218s 2 +0.0375s+0.0012 s 3 +0.4063s 2 +0.03227s+0.0001

⎤ ⎥ ⎥ ⎦

Table 4 Third-order transfer function relating F in1 and h3 Region

I

II

III

Transfer function between F in1 and h3 for each operating region

0.012 s 3 +1.295s 2 +0.3237s+0.0028

0.003 s 3 +0.65s 2 +0.0822s+0.0004

0.0012 s 3 +0.4063s 2 +0.0323s+0.0001

2.3 Transfer Function for the Third-Order Three Interacting Tank Process State space to transfer function conversion results in transfer function matrix as the system is a two-input three-output process. But the input of interest is the inflow to tank 1 and the output height of the third tank. The transfer function between height of third tank and the inflow to tank 1 is considered for this research which is a third-order system. For the three operating regions, the transfer function is found and tabulated in Table 4.

2.4 Model Order Reduction and Approximation to a Fractional-Order System Using Bee Colony Optimization Technique Model order reduction is required for reducing the computational complexity. In this section, the third-order system is converted to FOFOPDT system. As per Pade’s approximation, any higher-order process can be approximated to FOPDT system. Analysis and controller design are easier for a lower order. Fractionalizing in every

422

U. Sabura Banu et al.

domain is gaining popularity in the recent past. Here, in this work, an attempt is made to approximate the third-order process to FOFOPDT system. There is no straightforward approach to determine the system parameters of the approximated model. As swarm intelligence techniques are gaining popularity, bee colony optimization is used for determining the approximated model parameters.

3 Objective Function for the Approximation General form of FOFOPDT system is given by G FOFOPDT (s) =

K e−Ls τ sα + 1

(11)

where K τ α L

Gain of the FOFOPDT system. time constant. fractional order. time delay.

Optimizer requires minimization of the objective function and root mean square error (RMSE). Root mean square error (RMSE) is given by RMSE =

(Y3rd − YFOFOPDT )2 . N −1

(12)

where Y 3rd is the actual height of the third-order system and the Y FOFOPDT is the output of FOFOPDT system. The parameters K, τ, α and L are optimized to minimize the root mean square error.

4 Bee Colony Optimization for Model Approximation Bee colony optimization (BCO) is a swarm intelligence technique used for optimizing the FOFOPDT model parameters. In BCO, there are three groups of bees: employed bees, onlooker bees, and scout bees. Employed bees go in search of food source, and depending on the quantity of the food available in the source, they perform dance. The direction of the food source and the quantity of the honey are explicitly communicated by a dance by the recruiter or employed bees. The employed bees are managed by the onlooker or forager bees to collect the information regarding the quality of the food source and choose the best food source. The onlooker bees now move to that source. Scouts are used to replace the abandoned food source with new

FOPD Controller Using Bee Colony Optimized Reduced …

423

food sources. This sequence is represented as an artificial bee colony optimization algorithm. In the bee colony optimization algorithm, position of the food source is the parameter to be optimized and the quality and quantity of the nectar form represent the objective function or the fitness function. This optimization algorithm is an iteration-based and population-based. A number of bees are determined by the number of parameters to be optimized. The process starts with initial random population. Fitness of the function is computed for the entire population. Greedy selection process and probability function are used to get the parameters with best fit. Best solution is memorized. The worst solutions are updated to produce new solutions. The process is repeated till the stopping criteria is met.

5 Algorithm for Bee Colony Optimization for Model Approximation 1. Randomly initialize the population, x i,j 2. Determine the fitness of each individual in a population 3. Find the new solutions (food source positions) vi,j using the formula νi, j = xi, j + φi j (xi, j − xk, j )

(13)

where K and I are neighbors, φ is a random number. 4. Compute probability values Pi =

fiti SN

fiti

(14)

i=1

Objective function of the best fit is given by  fiti =

if f i ≥ 0 1 + abs( f i ) if f i < 0 1 1+ f i

 (15)

Normalize Pi values into [0, 1]. 5. Obtain new solutions vi for the onlookers from the solutions x i , for Pi and evaluate them 6. Apply the greedy selection process for the onlookers between vi and x i 7. Scouts are used to find the abandoned solution (source) and update it with new randomly produced solution x i for the scout using   xi j = min +rand(0, 1) ∗ max − min . j

j

j

(16)

424

U. Sabura Banu et al.

8. Memorize the best food source position (solution) achieved so far 9. Repeat steps 3–9 till stopping criteria met. Stopping criteria are either getting a minimum root mean square error or a given number of iterations. The number of iterations is selected as 100 in this work.

5.1 Approximation of the Third-Order System to FOFOPDT Using Bee Colony Optimization Figure 3 shows the block diagram for the online identification of FOFOPDT system using BCO. The third-order system and the identified FOFOPDT system are subjected to step input and the output deviation computed. With the error found, the root mean square error is computed. Table 5 shows the transfer function of the approximated models. To determine the modeling accuracy, the output responses of the third-order system and FOFOPDT system are analyzed in time domain and frequency domain.

Y3rd

Third Order Model for the three interacting tank process

+ Root Mean Square Error

Step Input

Fractional Order First Order Plus Dead Time Process

YFOFOPDT

Bee Colony Optimizer

Fig. 3 Block diagram of the FOFOPDT system approximation using BCO

Table 5 Original third-order transfer function, FOPDT system, and FOFOPDT system transfer function under different operating conditions Region

Third-order transfer function

FOPDT system transfer function

FOFOPDT system transfer function

I

0.012 s 3 +1.295s 2 +0.3237s+0.0028 0.003 s 3 +0.65s 2 +0.0822s+0.0004 0.0012 s 3 +0.4063s 2 +0.0323s+0.0001

4.286 −1.87s 120.344s+1 e 7.5 −3.46s 211.88s+1 e 12 −4.62s 343.619s+1 e

4.25 e−2.7s 138s 1.038 +1 7.455 e−8.6s 230.586s 1.0153 +1 11.5315 e−9.5534s 343.4865s 1.0238 +1

II III

FOPD Controller Using Bee Colony Optimized Reduced …

425

For time domain analysis, step input is applied for the third-order transfer function, FOPDT system, and fractional-order FOPDT system under different operating regions and performance compared in Table 6. The response shows that the output of all the three models matches both in the transient and steady-state region. Frequency response for the third-order process and the approximated process are shown in Table 6. Gain margin and phase margin also match. Table 6 shows the comparison of time domain and frequency domain of the third-order original process and the approximated model. Root mean square error determined between the third-order process and the approximated FOFOPDT process and the frequency domain specifications are reported in Table 7. Table 6 Time domain and frequency domain comparison I

II Time Domain Response

4

10

2 1.5 1

4 3

2

1 300

400 500 600 Time (in secs)

700

800

900

1000

0

700

800

Magnitude (dB)

-40

-60 0 -45 -90 -135 G3_3rd_order_TF

300

600 500 400 Time (in secs)

700

800

900

1000

Frequency Domain Response

-40

0 G3_3rd_order_TF

-20

G3_fopdt_TF G3_FO_fopdt_TF

-40

-60 0

-60 0

-180

-180

-360 G3_3rd_order_TF

-360 -540

G3_fopdt_TF G3_FO_fopdt_TF

-720

-720 -3

10

200

40

G3_FO_fopdt_TF -4

100

20

0

-20

-540

G3_fopdt_TF

-270 10

0

1000

900

20

0

-225

600 500 400 Time (in secs)

300

200

Frequency Domain Response

Frequency Domain Response 20

-20

-180

0

100

0

Magnitude (dB)

200

Phase (deg)

100

6

4

2

0

G3_3rd_order_TF G3_fopdt_TF G3_FO_fopdt_TF

8

G3_3rd_order_TF G3_fopdt_TF G3_FO_fopdt_TF

5

Height (in cm)

Height (in cm)

G3_3rd_order_TF G3_fopdt_TF G3_FO_fopdt_TF

2.5

0.5

Magnitude (dB)

12

7 6

3

Phase (deg)

Bode plot for the third-order system, FOPDT system, and fractional-order FOPDT system for the first region

Time Domain Response

8

4.5

3.5

0

III Time Domain Response

Phase (deg)

Comparison of the step response of the third-order system, FOPDT system, and fractional-order FOPDT system

Height (in cm)

Region

-1

-2

10

10

0

10

-4

10

-3

10

-2

10

-1

10

0

10

-4

10

Frequency (rad/sec)

Frequency (rad/sec)

-3

10

-2

10

-1

10

0

10

Frequency (rad/sec)

Table 7 Comparison of the frequency domain specification and root mean square error for the three interacting process at three different regions Region

System

Gm

Pm

Gain crossover frequency

Phase crossover frequency

First region

Third-order TF

34.70

94.8669

0.569

0.0371

Second region

Third region

RMSE Y 3rd − Y approx

FOPDT

23.58

99.78

0.84

0.0346

0.0433

FOFOPDT

25.6

94.3

0.596

0.0356

0.02691

Third-order TF

17.6794

80.9369

0.2867

0.0366

FOPDT

12.91

90.71

0.457

0.0351

0.1103

FOFOPDT

14.9

78.0

0.184

0.0356

0.1392

Third-order TF

10.8545

68.9708

0.1797

0.0361

85.57

0.34

0.0348

0.2646

69.9

0.146

0.0356

0.1363

FOPDT FOFOPDT

9.736 12.3

426

U. Sabura Banu et al.

6 Gain and Phase Computation for the FOFOPDT System For designing fractional-order controller, gain and the phase expressions need to be found. The transfer function of FOFOPTD is given by: P(s) =

K e−Ls ; τ sα + 1

(17)

K e− j Lω τ ( jω)α + 1

(18)

where K L τ α

gain of the process, delay of the system, time constant of the system. order of the integrator. Putting s = jω: P( jω) = Since  απ  απ s α = ( jω)α = ωα cos + j sin 2 2

(19)

So, substituting this value of Eqs. (2, 5) in above equation: P( jω) = ⇒ P( jω) =

K e− j Lω  τ ωα cos απ + j sin 2

απ 2



+1

− j Lω

Ke τ ωα cos απ + jτ ωα sin 2

απ 2

+1

(20)

Therefore, the magnitude of the above P(jω), will be given by: K |P( jω)| =  2  απ τ ωα cos 2 + 1 + τ ωα sin

 απ 2 2

(21)

And the phase angle is given by: Arg(P( jω)) = −Lω − tan−1



 τ ωα sin απ 2 . τ ωα cos απ +1 2

(22)

FOPD Controller Using Bee Colony Optimized Reduced …

427

7 Fractional-Order PD (FOPD) Controller The fractional-order PD controller parameters are designed by satisfying the phase margin specifications, gain crossover frequency, and robustness to variations in this section. For a FOFOPDT system (17), general representation of the fractional-order PD controller is represented as: C(s) = K p + K p K d s μ = K p (1 + K d s μ )

(23)

By substituting s = jω, in the above equation C( jω) = K p (1 + K d ( jω)μ )

(24)

Using trigonometrical representation j μ = cos

μπ μπ + j sin 2 2

(25)

Substituting this value j μ in Eq. (24), we get:   μπ  μπ + j sin C( jω) = K p (1 + K d ωμ ( j)μ )=K p 1 + K d ωμ cos 2 2 μπ μπ μ μ C( jω) =K p + K p K d ω cos + j K p K d ω sin (26) 2 2 Magnitude of C(jω) is represented by: 

μπ 2  μπ 2 K p + K p K d ωμ cos + K p K d ωμ sin 2 2  μπ |C( jω)| = K p 1 + K d2 ω2μ + 2K d ωμ cos 2 |C( jω)| =

(27)

And the phase angle of C(jω) is given by: Arg(C( jω)) = tan

−1



K d ωμ sin μπ 2 1 + K d ωμ cos μπ 2

 (28)

Normally, FOPD controller have three degrees of freedom and it is met by the three following specifications. (1) Phase margin specification, i.e.

428

U. Sabura Banu et al.

Arg[G( jωgc )] = Arg[C( jωgc )P( jωgc )] = −π + φm .

(29)

(2) Gain crossover frequency specification, i.e. |G( jωgc )| = |C( jωgc )P( jωgc )| = 1.

(30)

(3) Robustness to gain variation of the plant, i.e.      d(Arg(C( jω)P( jω)))   d(Arg(G( jω)))      = = 0.    dω dω ω=ωgc ω=ωgc

(31)

For phase margin specification from Eq. (29) we have: Arg[G( jωgc )] = Arg[C( jωgc )P( jωgc )] = −π + φm Substituting the phase and representing, we obtain  tan

−1

⇒ tan

μ sin K d ωgc μ

μπ 2



 − Lωgc − tan

1 + K d ωgc cos μπ 2  μπ μ ω sin K d gc 2 −1

−1

α cos τ ωgc



απ 2 απ + 2

+ π = (Lωgc + φm ) + tan

μ 1 + K d ωgc cos μπ 2



α sin τ ωgc

= −π + φm

1 

−1

απ 2 απ + 2

α sin τ ωgc α cos τ ωgc

 1 (32)

Taking tan on both sides, we have:  tan tan−1



μ K d ωgc sin μπ 2

μ 1 + K d ωgc cos μπ 2



 +π





= tan (Lωgc + φm ) + tan−1

α sin απ τ ωgc 2



α cos απ + 1 τ ωgc 2

(33)

Since L, ω and φm are known constant, so taking the LHS of Eq. (33) as X: 



X = tan (Lωgc + φm ) + tan

−1

απ 2 απ + 2



α sin τ ωgc α cos τ ωgc

1

(34)

Replacing this value of X in Eq. (33), we get:  tan tan

 −1

μ sin K d ωgc μ

μπ 2

1 + K d ωgc cos μπ 2



 +π

=X

(35)

FOPD Controller Using Bee Colony Optimized Reduced …

429

Simplifying above equation using tan(A + B) formula:    μ K ω sin μπ + tan π tan tan−1 1+Kd d ωgcgcμ cos2 μπ 2    =X μ μπ K ω sin 1 − tan tan−1 1+Kd d ωgcgcμ cos2 μπ . tan π

(36)

2

Substituting the value tanπ = 0, we obtain: μ K d ωgc sin μ

μπ 2

1 + K d ωgc cos μπ 2

=X

(37)

Solving for K d from above equation, finally we get: Kd =

μ ωgc sin μπ 2

X  − X cos μπ 2

(38)

Similarly for gain crossover frequency specification from Eq. (30) we have: |G( jωgc )| = |C( jωgc )P( jωgc )| = 1 Substituting the magnitude of C( jωgc ) and P( jωgc ) in above equation:  2μ μ K p K K d2 ωgc + 2K d ωgc cos μπ +1 2   =1   2 α sin απ 2 τ ωα cos απ + 1 + τ ω 2 2

(39)

Performing cross-multiplication, we get:  KpK

2μ μ K d2 ωgc + 2K d ωgc cos

μπ = 2



τ ωα cos

2  απ απ 2 + 1 + τ ωα sin 2 2 (40)

Solving for K p , we obtain:   2 2  τ ωα cos απ + 1 + τ ωα sin απ 1 2 2   Kp =   2μ μ K K d2 ωgc + 2K d ωgc cos μπ +1 2 Finally for robustness to gain variation of the plant from Eq. (31), we have:      d(Arg(C( jω)P( jω)))   d(Arg(G( jω)))      = =0    dω dω ω=ωgc ω=ωgc

(41)

430

U. Sabura Banu et al.

After substituting the values of phase of C(jω) and P(jω), we will have:    α απ    K ωμ sin μπ τ ω sin d tan−1 1+Kd d ωμ cos2 μπ − tan−1 τ ωα cos απ2+1 − Lω 2

2

dω    K ωμ sin μπ d tan−1 1+Kd d ωμ cos2 μπ 2





  α απ  τ ω sin d tan−1 τ ωα cos απ2+1 2





=0

d(Lω) =0 dω

(42)

(43)

After differentiating the above equation, we get:       1 + K d ωμ cos μπ . μK d ωμ−1 sin μπ − K d ωμ sin μπ . μK d ωμ−1 cos μπ 2 2 2 2   2   2 K d ωμ sin μπ 2 1 + 1+K d ωμ cos μπ . 1 + K d ωμ cos μπ 2 2  α    α   τ ω cos απ + 1 ατ ωα−1 sin απ − τ ω sin απ ατ ωα−1 cos απ 2 2 2 2   − −L=0 2   τ ωα sin απ 2 α cos απ + 1 2 + 1 τ ω τ ωα cos απ +1 2 2

(44) μK ωμ−1 sin

μπ

Simplifying 1+K 2 ω2μd +2K ωμ 2cos μπ − d d 2 At ω = ωgc , we have μπ 2 μ 2K d ωgc cos μπ 2

μ−1 sin μK d ωgc 2μ 1 + K d2 ωgc +

ατ ωα−1 sin απ 2 1+τ 2 ω2α +2τ ω cos

=

απ 2

− L = 0.

α−1 sin ατ ωgc

απ 2

2α s + 2τ ω cos απ 1 + τ 2 ωgc gc 2

+L

(45)

Taking the RHS of Eq. (45) as Z, we have: Z=

α−1 sin ατ ωgc

απ 2

2α + 2τ ω cos απ 1 + τ 2 ωgc gc 2

+L

(46)

Substituting the value of Eq. (46) in Eq. (50), we have: μπ 2 μ 2K d ωgc cos μπ 2

μ−1 μK d ωgc sin 2μ

1 + K d2 ωgc +

=Z

(47)

Simplifying the above equation  2μ Z K d2 ωgc

+ Z Kd

μ 2ωgc

μ−1 μωgc sin μπ − cos 2 Z

μπ 2

 +Z =0

(48)

FOPD Controller Using Bee Colony Optimized Reduced …

431

Now, substituting the value of K d from Eqs. (38)–(48), we obtain:

 sin

 μ μ−1 X 2Z ωgc cos μπ − μωgc sin 2   +  μ μπ μπ 2 ωgc sin 2 − X cos 2 − X cos μπ

2μ Z X 2 ωgc μπ 2

μπ 2

 +Z =0

(49)

2

Algebraic simplification leads to sin2

μπ 2



Xμ − X2Z − Z ωgc



 = sin μπ

 X 2μ . 2ωgc

(50)

From Eq. (50), the only tunable parameter is to be determined, i.e., μ. Equation (50) is solved graphically to have the value of μ by considering LHS as one function of μ i.e. F 1 (μ) and RHS as another function of μ i.e. F 2 (μ). Plot the graphs of F 1 (μ) and F 2 (μ) and find the intersection of both the curve to have the solution of μ. Also, keep in mind that select that value of μ such that it should lie between 0 and 2, i.e., μ ∈ (0, 2). μπ F1 (μ) = sin 2 2



Xμ − X2Z − Z ωgc

 (51)

and  F2 (μ) = sin μπ

X 2μ 2ωgc

 (52)

Once the value of μ is obtained, K d is obtained using Eq. (38). Similarly, the value of μ and K d is substituted in Eq. (41) to have the value of K p . These controller parameters are tabulated in Table8 for all the three regions. And also, the controllers transfer function for all the regions is tabulated in Table 9. The designed FOPD controller is implemented for the original system. For each region, closed-loop response is obtained which is tabulated in Table 10. A step input of 3 cm, 10 cm, and 20 cm is given for the first region, second region, and third region, respectively. It is obvious from the step responses of all the three regions that there is relatively large offset due to the absence of integral term. Therefore, these designed FOPD controllers are not feasible. The dip can be also observed in the closed-loop response, which is due to the right-hand-side zero.

8 Conclusion The designed FOPD controller is implemented for the original system. For each region, closed-loop response is obtained which is tabulated in the Table 9. It is obvious from the step responses of all the three regions that there is relatively large

432

U. Sabura Banu et al.

Table 8 FOPD controller parameters Graphical solution for μ for FOPD controller

Region

Design specifications

Controller parameters

1st region

ωgc = 0.1 (rad/s) ϕ m = 45°

FOPD μ = 1.755 K p = 1.1825 K d = 194

2nd region

ωgc = 0.05 (rad/s) ϕ m = 45°

FOPD μ = 1.841 K p = 1.02 K d = 593

3rd region

ωgc = 0.05 (rad/s) ϕ m = 30°

FOPD μ = 1.755 K p = 0.7699 K d = 516

Table 9 FOPD controller transfer function Regions

Third-order system

FOFOPDT system

Controller transfer function

1st region

0.012 s 3 +1.295s 2 +0.3237s+0.0028

4.25 e−2.7s 138s 1.038 +1

2nd region

0.003 s 3 +0.65s 2 +0.0822s+0.0004

7.455 e−8.6s 230.586s 1.0153 +1

3rd region

0.0012 s 3 +0.4063s 2 +0.0323s+0.0001

11.5315 e−9.5534s 343.4865s 1.0238 +1

C(s) = 1.1825+229.4s1.755 C(s) = 1.02 + 605s1.841 C(s) = 0.7699 + 397s1.755

Table 10 Step response for FOPD controllers First region

Second region

Third region

FOPD Controller Using Bee Colony Optimized Reduced …

433

offset due to the absence of integral term. Therefore, these designed FOPD controllers are not feasible. The dip can be also observed in the closed-loop response, which is due to the right-hand-side poles. The performance can be improved by design of FOPI controller.

References 1. Oldham KB, Spanier J (2006) The fractional calculus: theory and applications of differentiation and integration to arbitrary order. Dover Books on Mathematics 2. Lou Y, Chen YQ, Wang CY, Pi YG (2010) Tuning fractional order proportional integral controllers for fractional order systems. J Process Control 20:823–832 3. Hamamci SE, Kolsal M (2010) Calculation of all stabilizing fractional order PD controllers for Integrating time delay systems. Comput Math Appl 59(5):1630–1636 4. Wang DJ (2012) H∞ design with fractional order PDμ controller. Automatica 48(5):974–977 5. Hamamci SE (2007) An algorithm for stabilization of fractional order time delay systems using fractional order PID controllers. IEEE Trans Autom Control 52(10):1964–1969 6. Hamamci SE (2008) Stabilization using fractional order PI and PID controllers. Nonlinear Dyn 51:329–343 7. Chen YQ, Bhaskaran T, Xue D (2008) Practical tuning rule development for fractional order proportional and integral controllers. ASME J Comput Nonlinear Dyn Syst 3(2):021403 8. Monje CA, Calderon AJ, Vingre BM, Chen YQ, Feliu F (2004) On fractional PIλ controllers; some tuning rules for robustness to plant uncertainties. Nonlinear Dyn 38(1–2):369–381 9. Li HS, Luo Y, Chen YQ (2010) A fractional order proportional and derivative (FOPD) motion controller: tuning rule and experiments. IEEE Trans Control Syst Technol 18(2):516–520 10. FabrizioPadula AV (2011) Tuning rules for optimal PID and fractional-order PID controllers. J Process Control 21:69–81 11. Özbay H, Bonnetb C, Fioravanti AR (2012) PID controller design for fractional-order systems with time delays. Syst Control Letters 61:18–23 12. Moghaddam TV, Bigdeli N, Afshar K (2011) Tuning a fractional order PID controller with lead compensator in frequency domain. World Acad Sci Eng Technol 51:1277–1282 13. Wang D-J, Li W, Guo M-L (2013) Tuning of PID controllers based on sensitivity constraint. J Process Control 23:861–867 14. Hajiloo A, Nariman-Zadeh N, Moeini A (2012) Pareto optimal robust design of fractional-order PID controllers for systems with probabilistic uncertainties. Mechatronics 22:788–801 15. Das S, Saha S, Das S, Gupta A (2011) On the selection of tuning methodology of FOPID controllers for the control of higher order processes . ISA Trans 50:376–388 16. Saha S, Das S, Das S, Gupta A (2012) A conformal mapping based fractional order approach for sub-optimal tuning of PID controllers with guaranteed dominant pole placement. Commun Nonlinear Sci Numer Simul 17:3628–3642 17. Pan I, Das S (2013) Frequency domain design of fractional order PID controller for AVR system using chaotic multi-objective optimization. Electr Power Energy Syst 51:106–118 18. Das S, Pan I, Das S, Gupta A (2012) A novel fractional order fuzzy PID controller and its optimal time domain tuning based on integral performance indices. Eng Appl Artif Intell 25:430–442 19. Nakrani S, Tovey C (2004) On honey bees and dynamic allocation in an internet server colony. Adaptive Behavior 12(3–4):223–240 20. Teodorovic D, Dell’orco M (2005) Bee colony optimization—a cooperative learning approach to complex transportation problems. Adv OR AI Methods Transp: 51–60 21. Seeley TD The wisdom of the hive. Publication Harward University Press 22. Lakshmanaprabu SK, Sabura Banu U (2018) Real time implementation of multivariable centralised FOPID controller for TITO process. Int J Reason Based Intell Syst Artif Intell Tech Appl Study Eng Appl Part I 10(3/4):242–251

434

U. Sabura Banu et al.

23. Lakshmanaprabu SK, Wahid Nasir A, Sabura Banu U (2017) Design of centralized fractional order PI controller for two interacting conical frustum tank level process. J Appl Fluid Mech 10:23–32. ISSN 1735-3572, EISSN 1735-3645 24. Lakshmanaprabu SK, Sabura Banu U (2017) Adaptive multiloop IMC based PID controller tuning using Bat optimization algorithm for two interacting conical tank process. Int J Adv Intell Parad 25. Sabura Banu U, Wahid Nasir A (2015) Design, analysis and performance evaluation of fractional order proportional integral controller for three interacting tank process in frequency domain considered as third order system. Int Feder Autom Control IFAC-Papers OnLine Sci Direct 48(30):179–218

Hardware Security for DSP Circuits Using Key-Based Time-Dependent Obfuscation R. Uma Rani, D. Jayanthi, Sudharsan Jayabalan, and Arun Vignesh

Abstract Integrated circuit (IC) design flow undergoes multistep process concentrating on major factors like power, time, and area. By side of this, hardware security became major consideration factor due to increase in hardware design vulnerabilities. Most of the vendors use intellectual property (IP) cores and proceed with further design process which includes some suspicious activities during design flow making designers to ensure design trust by reevaluating. This increased focus on suspicious activities which leads to attacks such as reverse engineering, illegal use of IC/IP, and illegal IC overbuilding. Obfuscation can prevent such activities by obfuscating the IP/IC functionality using concept of secret keys. This paper proposes advance technique of random natured time-dependent obfuscation for sequential circuits. This method protects the design functionality controlled by control signals along with datapath obfuscation, which increases circuit security level compared to existing obfuscation methods. This concept is applicable to any DSP applications. It ensures high degree of security with low power and area overheads. Keywords Obfuscation · Reverse engineering · Suspicious activities · Time-dependent obfuscation

1 Introduction Advances in IC manufacturing with multiple capabilities and emerging technologies lead to increase in the cost of owning a foundry with mixed technologies [1, 2] which cannot be afforded by design houses and hence uses third-party IP cores which is fabricated in a foundry not under the control of design house and leads to issues explained by Rostami et al. with affect semiconductor manufacturing industries revenue [3]. Rajendran et al. elucidated the vulnerabilities in IC design, fabrication process, and about techniques which retain IC trust [4], which is classified as logic encryption, split-manufacturing, IC camouflaging, and Trojan activation. R. Uma Rani · D. Jayanthi (B) · S. Jayabalan · A. Vignesh Department of Electronics and Communication Engineering, Gokaraju Rangaraju Institute of Engineering and Technology, Hyderabad, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 P. Siano and K. Jamuna (eds.), Advances in Smart Grid Technology, Lecture Notes in Electrical Engineering 687, https://doi.org/10.1007/978-981-15-7245-6_32

435

436

R. Uma Rani et al.

Logic encryption is a locking concept on ICs/IPs to avoid reverse engineering and illegal overbuilding by third parties and obstruct insertion of harmful Trojan logic by attackers. Subramanyan et al. illustrated attacks on logic encryption and algorithm like SAT-based algorithm determines consistent of key values, and partial-break algorithm determines correct value for some key inputs when some of its values are unique in nature, where fault analysis algorithm can decrypt only few key values [5]. To prevent malicious logic insertion, Dupuis et al. shed light on logic encryption technique with objective of minimizing number of low controllability signals [6] which are used for Trojan insertion by attacker. K. Juretus et al. proposed transmission gate and stack-based topology for encryption [7] to reduce overheads due to XOR or LUTs logic encryption techniques. Further research on securing hardware design is moved to a technique where instead of additional logic encryption gates, the circuit functionality is camouflaged by structural modifications. Chakraborty et al. developed technique to protect and authenticate hardware through netlist-level obfuscation method, and this concept is illustrated by considering FSM. Here, the structural modification is done by selecting specified internal nodes and state transition function. Maximum functional, structural obfuscation with minimal overhead is obtained by modifying pre-synthesized IP core gate-level netlist followed by re-synthesis [8, 9]. IP cores come in the form of RTL description, i.e., soft IP or gate-level description, i.e., firm IP or GDS-II design data ,i.e., hard IP, and these IPs are highly unguarded to illegal reuse issues by entrusted third party. Narasimhan et al. illustrated a solution for this issue by embedding FSM as a Trojan circuit inside an IP core which reduces life span of IP by interrupting normal function when it detects any occurrence of rare events [10]. Currently, FPGAs have become a popular design with low design cost, reprogrammability, and target for IP dereliction. To overcome this, Lin et al. developed IP protect technique by restricting IPs execution only on specified FPGAs. Here, FPGAs are customized by PUFs and FSMs are embedded in the IP such that FSM gets activated by PUF response from FPGA [11]. Similar to logic encryption, logic obfuscation technique is illustrated by Zhang et al. to prevent reverse engineering of gate-level netlist, layout-level geometry of IP/IC, and about combinational obfuscation. He introduced concept of obfuscation cell, which is a combination of inverter and multiplexer and provides resistance against brute-force attacks [12]. Ayinala et al. derived concept from class of high-level transformations to secure hardware design, and the work was illustrated on FFT circuit employing folding transformation [13, 14] Garrido et al. implemented this concept on real-valued signals. Later, Shanmugam et al. worked on FFT architecture with obfuscation technique by introducing modes concept in control path, which gives correct output in desired mode and partial or incorrect output in undesired mode [15]. Same FFT architecture is illustrated by adopting high-level transformations in the design by Lao et al., here the folded architecture of FFT circuit implemented which can be obfuscated based on the control signals is derived from the control path [16], and this concept is also demonstrated with fast Fourier transform circuits by Sekhar et al. [17]. Kim et al. [18] proposed a method to obfuscate the design functionality by changing control path instead of varying data path. In this method, different modes

Hardware Security for DSP Circuits Using Key-Based …

437

are derived through control path modifications and type of mode is defined by key bits. Xie et al. demonstrated the drawbacks of fixed obfuscation method in terms of reduced security level and susceptibility (SAT)-based attacks on circuits performed by analysis circuit parameters [19]. Yasin et al. [1] implemented a logic technique which can distinguish maximum number of required input patterns to recapture the secret keys, here a small comparator circuit which generates a signal is activated only for particular input and key bits, and this comparator is used to construct a SAT attack resistant circuit. Koteshwara et al. elaborated this concept by introducing concept of key-based dynamic obfuscation concept, where the corruption of output signals is in random nature and predictability time of correct key and circuit security level is increased [20]. In this paper, the obfuscation concept is further elaborated with high degree of security level by obfuscating design at both data path and control path. Here, the mixed concept of obfuscation method is introduced to achieve high-level protection of circuit functionality and low overheads, and this concept can be implemented to any critical DSP or other applications.

2 Architecture Illustration The proposed obfuscated architecture shown in Fig. 1 applicable to any sequential circuits is illustrated in this paper. Here, it is examined on DSP application, i.e., folded FFT algorithm and it can be implemented on any critical applications. The block diagram represents data path, mode selector shown in Fig. 2, control path and control signals obfuscation circuit shown in Fig. 3, and circuit for hamming distance computation. This method is derived from existing obfuscation method termed as mixed obfuscation and obfuscates circuit architecture, internal signals by generating corrupted output values for incorrect key sequences.

2.1 Datapath Architecture Figure 2 represents folded, Radix-22 , 1024/256 DIF FFT architecture controlled by control and mode bits. The basic elements are represented by BFI and BFII which perform real arithmetic operations and delay switches, mode block baffles internal signals of data path by modifying actual architecture of the DSP design; any combination of architectures, i.e., 64, 32, 16, 8 can be implemented by selecting mode block and delay switches based on circuit requirement. Mode block: Based on mode bits m0, m1, respective internal signal gets selected which specifies datapath architecture. Let us consider data path with architectures 1024/256, here, x1 and x2 represents signals for next delay element where x1