Neural and Fuzzy Logic Control of Drives and Power Systems Neural and Fuzzy Logic Control of Drives and Power Systems M.N. Cirstea, A. Dinu, J.G. Khor, M. McCormick Newnes OXFORD AMSTERDAM BOSTON LONDON NEW YORK PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO Newnes An imprint of Elsevier Science Linacre House, Jordan Hill, Oxford OX2 8DP 225 Wildwood Avenue, Woburn, MA 01801-2041 First published 2002 Copyright © 2002, M.N. Cirstea, A. Dinu, J.G. Khor, M. McCormick. All rights reserved The right of M.N. Cirstea, A. Dinu, J.G. Khor and M. McCormick to be identified as the authors of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 No part of this publication may be reproduced in any material form (including photocopying or storing in any medium by electronic means and whether or not transiently or incidentally to some other use of this publication) without the written permission of the copyright holder except in accordance with the provisions of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London, England W1T 4LP. Applications for the copyright holder’s written permission to reproduce any part of this publication should be addressed to the publisher British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 0 7506 55585 For information on all Newnes publications visit our website at www.newnespress.com Typeset at Replika Press Pvt Ltd, Delhi 110 040, India Printed and bound in Great Britain Preface Control systems 1 Control theory: historical review 1 Introduction to control systems 2 Control systems for a. c. drives 5 Modern control systems design using CAD techniques Electronic design automation ( EDA) Application specific integrated circuit ( ASIC) basics 12 Field programmable gate arrays ( FPGAs) 14 ASICs for power systems and drives 16 Electric motors and power systems Electric motors Power systems 19 Pulse width modulation 22 The space vector in electrical systems 26 Induction motor control 28 Synchronous generators control 51 Elements of neural control Neurone types Artificial neural networks architectures 59 Training algorithms 61 Control applications of ANNs 69 Neural network implementation 71 Neural FPGA implementation Neural networks design and implementation strategy Universal programs  FFANN hardware implementation 95 Hardware implementation complexity analysis 98 Fuzzy logic fundamentals Historical review Fuzzy sets and fuzzy logic 114 Types of membership functions 116 Linguistic variables 117 Fuzzy logic operators 117 Fuzzy control systems 118 Fuzzy logic in power and control applications 121 VHDL fundamentals Introduction VHDL design units 126 Libraries, visibility and state system in VHDL 131 Sequential statements 135 Concurrent statements 141 Functions and procedures 146 Advanced features in VHDL 151 Summary 154 Neural current and speed control of induction motors The induction motor equivalent circuit The current control algorithm 161 The new sensorless motor control strategy 183 Induction motor controller VHDL design 199 FPGA controller experimental results 227 Fuzzy logic control of a synchronous generator set System representation VHDL modelling 248 FPGA implementation 270 System assembly and experimental tests 285 Conclusions 292 Final notes References Appendices Appendix A - C++ code for ANN implementation Appendix B - C++ Programs for PWM generation 333 Appendix C - Subnetworks VHDL models 341 Appendix D - VHDL model of sine wave ROM 355 Appendix E - VHDL code for simulation 357 Appendix F - VHDL code for synthesis 374 Appendix G - PWM controllers 389 Index Preface The idea of writing this book arose from the need to investigate the main principles of modern power electronic control strategies, using fuzzy logic and neural networks, for research and teaching. Primarily, the book aims to be a quick learning guide for postgraduate/undergraduate students or design engineers interested in learning the fundamentals of modern control of drives and power systems in conjunction with the powerful design methodology based on VHDL. At the same time, the book is structured to address the more complex needs of professional designers, using VHDL for neural and fuzzy logic systems design, by including comprehensive design examples. This facilitates the understanding of hardware description language applications and provides a practical approach to the development of advanced controllers for power electronics. The first section of the book contains a brief review of control strategies for electric drives/power systems and a summary description of neural networks, fuzzy logic, electronic design automation (EDA) techniques, ASICs/FPGAs and VHDL. The aspects covered allow a basic understanding of the main principles of modern control. The second section contains two comprehensive case studies. The first deals with neural current and speed control of induction motor drives, whereas the second presents the environmentally friendly fuzzy logic control of a diesel-driven stand-alone synchronous generator set. Both control strategies were implemented in Xilinx FPGAs and comprehensively tested by simulation and experimental measurements. This book brings together the complex features of control strategies, EDA, neural networks, fuzzy logic, electric machines and drives, power systems and VHDL and forms a basic guide for the understanding of the fundamental principles of modern power electronic control systems design. To be expert in the design of advanced digital controllers for drives and power systems, extra reading is strongly recommended and comprehensive material is referenced in the bibliographical section. The book includes a number of recent research results from work carried out by the authors, who are members of the electronic control and drives research group at De Montfort University, Leicester, UK. The facilities provided by the university and the support of NEWAGE AVK SEG, Stamford, UK, a major international manufacturer of electric generators, are gratefully acknowledged. Dr Marcian N. Cirstea Dr Andrei Dinu Dr Jeen G. Khor Prof. Malcolm McCormick 1 Control systems 1.1 Control theory: historical review The function of a control mechanism is to maintain certain essential properties of a system at a desired value under perturbations. Historical control systems which are simple but effective have been employed in water regulation and control of liquid level in wine vessels for centuries. Some of these concepts are still used today, for example the float system in the water tank of the toilet flush. However, modern control systems used in today’s industry are much more complex and owe their beginnings to the development of control theory. The earliest significant work in modern automatic control can be traced to James Watt’s design of the fly-ball governor (1788) for the speed control of a steam engine. In 1868, Maxwell [170] presented the first mathematical analysis of feedback control. It was during this time that systematic studies into control systems and feedback dynamics began. One significant development was the well- known Routh’s stability criterion (1877) which won E.J. Routh the Adam’s Prize. The early twentieth century saw the beginning of what is now known as classical control theory. Minorsky’s work (1922) on the determination of stability from the differential equation describing the system (characteristic equation) and Nyquist’s development (1932) of a graphical procedure for determining stability (frequency response) substantially contributed to the study of control theory. In 1934, Hazen [111] introduced the term ‘servomechanism’ to describe position control systems in his attempt to develop a generalised theory of servomechanisms. Two years later, the development of the proportional integral derivative (PID) controller was described by Callender et al. (1936). Control theory, like many branches of engineering, underwent significant development during World War II. Based on Nyquist’s work, H.W. Bode introduced a method for feedback amplifier design, now known as the Bode plot (1945). By 1948, the root locus method of design and stability analysis was developed by W.R. Evans [93]. With the introduction of digital computers in the 1960s, the use of frequency response and characteristic equations began to give way to ordinary differential equations (ODEs), which worked well with computers. This led to the birth of modern control theory. While the term classical control theory is used to describe the design methods of Bode, Nyquist, Minorsky and similar workers, modern control theory relies on ODE design methods that are more suitable for computer aided engineering, for example the state space approach. Both these branches of control theory rely on mathematical representation of the control plant from which to derive its performance. To address the issues of non-linearities and time-variant parameters in plant models, control strategies [...]... the serial or parallel interfaces of a PC 16 Neural and Fuzzy Logic Control of Drives and Power Systems 2.4 ASICs for power systems and drives The development of a traditional microprocessor-based motion control system is a complex task consisting of several stages usually completed by several engineers It involves the design of both hardware and software components and their integration considering... open-loop control systems 4 Neural and Fuzzy Logic Control of Drives and Power Systems (E) The method of generating the control pulses: • Single-channel control systems • Multi-channel control systems (F) The synchronisation between the signals within the control system and input voltages: • Synchronous control systems • Asynchronous control systems 1.2.2 Characteristics of control systems Although different... reductions and performance improvements for a.c drive systems to make them more universally used Some of the expanding application areas are: • Replacement of variable speed d.c drives by appropriate a.c drive systems • Application of adjustable speed a.c drives to constant speed process control, thereby saving energy 6 Neural and Fuzzy Logic Control of Drives and Power Systems • Replacement of heat... two components into a single integrated circuit, the power integrated circuit (PIC) A PIC is defined by Thomas [217] as an integrated circuit which combines the logic level control and/ or protection circuitry with power handling capability of supplying 1 A and withstanding at least 100 V With the current 22 Neural and Fuzzy Logic Control of Drives and Power Systems trend towards integrated solutions,...2 Neural and Fuzzy Logic Control of Drives and Power Systems that continuously adapt to the variations of plant characteristics have been introduced Generally known as adaptive control systems, they include techniques such as selftuning control, H-infinity control, model referencing adaptive control and sliding mode control, Studies also include the use of non-linear state observers... Table 3.1 PWM control T1 Vcontrol ≥ Vtri Vcontrol < Vtri T2 T3 T4 ON OFF OFF ON OFF ON ON OFF Electric motors and power systems 23 In sinusoidal-PWM control schemes, there are two characteristic ratios which are important factors in the design of the controllers The amplitude modulation ratio ma is defined as the ratio of the peak amplitude of the control signal to the peak amplitude of the carrier... advantages of PLDs compared to FPGAs are the speed and ease of use without non-recurring engineering cost The size of PLDs is, on the other hand, smaller than that of FPGAs Current PLDs offer complexity equivalent to hundreds of thousands of gates and speed of the order of hundreds of MHz 2.3 Field programmable gate arrays (FPGAs) Field programmable gate arrays (FPGAs) are a special class of ASICs which... which provides position and velocity control for d.c., d.c brushless and stepper motors The HCTL-1000 executes any one of four control algorithms selected by the user: position control, proportional velocity control, trapezoidal profile control for point-topoint moves and integral velocity control The Signetics HEF4752V a.c motor control circuit is an ASIC designed for the control of three-phase pulse... It was the introduction of the glass bulb mercury arc rectifier (1900) which led to the beginning of the power electronics era Power electronics is the branch of engineering concerned with the application of electronics in the control and conversion of electrical power Early power electronic devices such as thyratrons and ignitrons were crude and unreliable The introduction of selenium rectifiers during... overview on control systems and their general features aimed to familiarise the reader with basic characteristics of control systems The next section focuses on some general aspects of control systems for electrical drives, especially for a.c electrical drives 1.3 Control systems for a.c drives A specific definition of a process control system may be: ‘A control system is a combination of amplifiers, . Neural and Fuzzy Logic Control of Drives and Power Systems Neural and Fuzzy Logic Control of Drives and Power Systems M. N. Cirstea, A. Dinu, J.G. Khor, M. McCormick Newnes OXFORD AMSTERDAM. and Fuzzy Logic Control of Drives and Power Systems (E) The method of generating the control pulses: • Single-channel control systems. • Multi-channel control systems. (F) The synchronisation. neural control Neurone types Artificial neural networks architectures 59 Training algorithms 61 Control applications of ANNs 69 Neural network implementation 71 Neural FPGA implementation Neural