Neural network applications in control

312 Pages·1995·14.118 MB·English

by  Irwin, G. W., Warwick, K., Hunt, K. J

CONTROL ENGINEERING SERIES 53
NEURAL
NETWORK
APPLICATIONS
IN CONTROL
Edited by
G. W. Irwin,
K. Warwick
and K. J. Hunt
The Institution of Electrical Engineers
IEE CONTROL ENGINEERING SERIES 53
Series Editors: Professor D. P. Atherton
Professor G. I. Irwin
NEURAL
NETWORK
APPLICATIONS
IN CONTROL
Other volumes In this series:
Volume 1 Multlvarlable control theory J. M. Layton
Volume 2 Elevator traffic analysis, design and control G. C. Barney and S. M. dos Santos
Volume 3 Transducers in digital systems Q. A. Woolvet
Volume 4 Supervisory remote control systems R. E. Young
Volume 5 Structure of interconnected systems H. Nicholson
Volume 6 Power system control M. J. H. Sterling
Volume 7 Feedback and multlvarlable systems D. H. Owens
Volume 8 A history of control engineering, 1800-1930 S. Bennett
Volume 9 Modern approaches to control system design N. Munro (Editor)
Volume 10 Control of time delay systems J. E. Marshall
Volume 11 Biological systems, modelling and control D. A. Linkens
Volume 12 Modelling of dynamical systems—I H. Nicholson (Editor)
Volume 13 Modelling of dynamical systems—2 H. Nicholson (Editor)
Volume 14 Optimal relay and saturating control system synthesis E. P. Ryan
Volume 15 Self-tuning and adaptive control: theory and application
C. J. Harris and S. A. Billings (Editors)
Volume 16 Systems modelling and optimisation P. Nash
Volume 17 Control in hazardous environments R. E. Young
Volume 18 Applied control theory J. R. Leigh
Volume 19 Stepping motors: a guide to modern theory and practice P. P. Acarnley
Volume 20 Design of modern control systems D. J. Bell, P. A. Cook and N. Munro (Editors)
Volume 21 Computer control of industrial processes S. Bennett and D. A. Linkens (Editors)
Volume 22 Digital signal processing N. B. Jones (Editor)
Volume 23 Robotic technology A. Pugh (Editor)
Volume 24 Real-time computer control S. Bennett and D. A. Linkens (Editors)
Volume 25 Nonlinear system design S. A. Billings, J. O. Gray and D. H. Owens (Editors)
Volume 26 Measurement and instrumentation for control M. G. Mylroi and G. Calvert
(Editors)
Volume 27 Process dynamics estimation and control A. Johnson
Volume 28 Robots and automated manufacture J. Billingsley (Editor)
Volume 29 Industrial digital control systems K. Warwick and D. Rees (Editors)
Volume 30 Electromagnetic suspension—dynamics and control P. K. Slnha
Volume 31 Modelling and control of fermentation processes J. R. Leigh (Editor)
Volume 32 Multivariable control for industrial applications J. O'Reilly (Editor)
Volume 33 Temperature measurement and control J. R. Leigh
Volume 34 Singular perturbation methodology In control systems D. S. Naidu
Volume 35 Implementation of self-tuning controllers K. Warwick (Editor)
Volume 36 Robot control K. Warwick and A. Pugh (Editors)
Volume 37 Industrial digital control systems (revised edition) K. Warwick and D. Rees
(Editors)
Volume 38 Parallel processing in control P. J. Fleming (Editor)
Volume 39 Continuous time controller design R. Balasubramanian
Volume 40 Deterministic control of uncertain systems A. S. I. Zinober (Editor)
Volume 41 Computer control of real-time processes S. Bennett and G. S. Virk (Editors)
Volume 42 Digital signal processing: principles, devices and applications
N. B. Jones and J. D. McK. Watson (Editors)
Volume 43 Trends in Information technology D. A. Linkens and R. I. Nicolson (Editors)
Volume 44 Knowledge-based systems for industrial control J. McGhee, M. J. Grimble and
A. Mowforth (Editors)
Volume 45 Control theory—a guided tour J. R. Leigh
Volume 46 Neural networks for control and systems K. Warwick, G. W. Irwin and K. J. Hunt
(Editors)
Volume 47 A history of control engineering, 1930-1956 S. Bennett
Volume 48 MATLAB toolboxes and applications for control A. J. Chipperfield and
P. J. Fleming (Editors)
Volume 49 Polynomial methods in optimal control and filtering K. J. Hunt (Editor)
Volume 50 Programming industrial control systems using IEC 1131-3 R. W. Lewis
Volume 51 Advanced robotics and intelligent machines J. O. Gray and D. G. Caldwell
(Editors)
Volume 52 Adaptive prediction and predictive control P. P. Kanjilal
NEURAL
NETWORK
APPLICATIONS
IN CONTROL
Edited by
G. W. Irwin,
K. Warwick
and K.). Hunt
The Institution of Electrical Engineers
Published by: The Institution of Electrical Engineers, London,
United Kingdom
© 1995: The Institution of Electrical Engineers
This publication is copyright under the Berne Convention and the
Universal Copyright Convention. All rights reserved. Apart from any fair
dealing for the purposes of research or private study, or criticism or
review, as permitted under the Copyright, Designs and Patents Act, 1988,
this publication may be reproduced, stored or transmitted, in any forms or
by any means, only with the prior permission in writing of the publishers,
or in the case of reprographic reproduction in accordance with the terms
of licences issued by the Copyright Licensing Agency. Inquiries
concerning reproduction outside those terms should be sent to the
publishers at the undermentioned address:
The Institution of Electrical Engineers,
Michael Faraday House,
Six Hills Way, Stevenage,
Herts. SG1 2AY, United Kingdom
While the editors and the publishers believe that the information and
guidance given in this work is correct, all parties must rely upon their own
skill and judgment when making use of it. Neither the editors nor the
publishers assume any liability to anyone for any loss or damage caused
by any error or omission in the work, whether such error or omission is
the result of negligence or any other cause. Any and all such liability is
disclaimed.
The moral right of the authors to be identified as authors of this work has
been asserted by them in accordance with the Copyright, Designs and
Patents Act 1988.
British Library Cataloguing in Publication Data
A CIP catalogue record for this book
is available from the British Library
ISBN 0 85296 852 3
Contents
Page
Preface xi
Contributors xiii
1 Neural networks: an introduction 1 arwick 1
1.1 Introduction 1
1.2 Neural network principles 2
1.3 Neural network elements 3
1.4 Hopfield networks 4
1.5 Kohonen networks 4
1.6 Multi-layer perceptrons 5
1.7 Radial basis function networks 9
1.8 N-tuple networks 13
1.9 Conclusions 14
1.10 References 15
Digital neural networks A. Redgers and I. Aleksander 17
2.1 Classification of neural networks 17
2.1.1 McCulloch and Pitts versus Boolean nodes 17
2.1.2 Definition of ANNs 18
2.1.3 Limitations of the classification scheme 20
2.2 Functions of Boolean nodes 20
2.2.1 MPLN activation functions 20
2.2.2 The address decoder function 21
2.2.3 Kernel functions and transforming output functions 22
2.3 Large logic neurons 23
2.3.1 Decomposition of Boolean nodes 23
2.3.2 Generalisation and generalising RAMs 24
2.3.3 Discriminators 25
2.3.4 Reversing discriminators 26
2.3.5 Pyramids 27
2.3.6 Networks of LLNs 28
2.4 Implementing large ANNs using MAGNUS 29
vi Neural network applications in control
2.4.1 General neural units 29
2.4.2 MAGNUS and LLNs 29
2.5 Conclusion 30
2.6 References 31
Fundamentals of neurocontrol: a survey 33
R, Zbikowski and P.J. Gawthrop
3.1 Introduction 33
3.2 Basic notions 34
3.3 Feedforward networks 35
3.3.1 Feedforward networks' description 35
3.3.2 Approximation: feedforward networks 37
3.3.3 Stone-Weierstrass theorem 38
3.3.4 Kolmogorov's theorem 40
3.3.5 Multi-dimensional sampling 42
3.3.6 Backpropagation for feedforward networks 44
3.4 Recurrent networks 46
3.4.1 Mathematical description 46
3.4.2 Approximation: recurrent networks 48
3.4.3 Fixed-point learning 50
3.4.4 Trajectory learning 51
3.4.5 Some control aspects of recurrent networks 57
3.4.6 Stability of recurrent networks 58
3.5 Conclusions 59
3.6 Acknowledgments 60
3.7 References 60
Selection of neural network structures: some approximation
theory guidelines J.C. Mason and P.C. Parks 67
4.1 Introduction: artificial neural networks (ANNs) in control 67
4.2 Approximation of functions 68
4.2.1 Choices of norm 71
4.2.2 Choice of form 72
4.2.3 Existence of good approximations 77
4.2.4 Approximation by interpolation 79
4.2.5 Existence and uniqueness of best approximation 80
4.2.6 Parallel algorithms 80
4.2.7 Training/learning procedures and approximation
algorithms 81
4.2.8 Practicalities of approximation 84
4.3 Convergence of weight training algorithms 85
4.4 Conclusions 85
4.5 References 87
Electric power and chemical process applications 91
G.W. Irwin, P. O'Reilly, G. Lightbody, M Brown
and E. Swidenbank
5.1 Introduction 91
Contents vii
5.2 Modelling of a 200 MW boiler system 91
5.2.1 Identification of ARMAX models 92
5.2.2 Neural boiler modelling: local models 94
5.2.3 Neural boiler modelling: global model 97
5.3 Inferential estimation of viscosity in a chemical process 98
5.3.1 The polymerisation process and viscosity control 98
5.3.2 Viscosity prediction using feedforward
neural networks 100
5.3.3 Viscosity prediction using B-spline networks 102
5.4 Acknowledgment 107
5.5 References 107
Studies in artificial neural network based control 109
KJ. Hunt and D. Sbarbaro
6.1 Introduction 109
6.2 Representation and identification 109
6.2.1 Networks, approximation and modelling 110
6.2.2 Identification 113
6.2.3 Modelling the disturbances 118
6.3 Gaussian networks 119
6.4 Learning algorithms 120
6.4.1 Modelling the plant: series-parallel model 120
6.4.2 Inverse model identification 121
6.5 Control structures 123
6.5.1 Supervised control 123
6.5.2 Direct inverse control 124
6.5.3 Model reference control 124
6.5.4 Internal model control 125
6.5.5 Predictive control 128
6.6 Example: pH control 133
6.7 Acknowledgments 135
6.8 References 135
Applications of dynamic artificial neural networks in state
estimation and nonlinear process control 141
P. Turner, J. Morris and G. Montague
7.1 Abstract 141
7.2 Introduction 142
7.3 Dynamic neural networks 142
7.4 Assessment of model validity 144
7.5 Data collection and pre-processing 145
7.6 Neural networks in state estimation 146
7.6.1 Estimation of polymer quality in an industrial
continuous polymerisation reactor 147
7.7 Neural networks in feedback process control 150
7.7.1 Inferential control 150
7.7.2 Auto-tuning feedback control 151
7.8 Neural networks in nonlinear model based control 151
viii Neural network applications in control
7.8.1 Application to an industrial high purity
distillation tower 154
7.9 Concluding remarks 155
7.10 Acknowledgments " 157
7.11 References 157
8 Speech, vision and colour applications 161
RJ. Mitchell and JM. Bishop
8.1 Introduction 161
8.2 Neural network based vision system 161
8.2.1 N-tuple vision system 162
8.2.2 A practical vision system 163
8.2.3 Implementation 165
8.2.4 Performance 166
8.2.5 Conclusion 166
8.3 A hybrid neural network system 166
8.3.1 Stochastic diffusion networks 167
8.3.2 Using analogue information 168
8.3.3 The hybrid system 168
8.4 The application of neural networks to computer
recipe prediction 169
8.4.1 Introduction 169
8.4.2 Computer recipe prediction 171
8.4.3 Neural networks and recipe prediction 171
8.4.4 Discussion 173
8.5 Kohonen networks for Chinese speech recognition 173
8.5.1 Review of Kohonen networks 173
8.5.2 Use in speech recognition 174
8.5.3 Chinese phoneme characteristics 175
8.5.4 Neural network system 175
8.6 Conclusion 176
8.7 References 176
9 Real-time drive control with neural networks 179
D. Neumerkel, J. Franz and L Kruger
9.1 Introduction 179
9.2 Induction machine control concepts 180
9.3 Model predictive control 183
9.3.1 Interval search optimisation 186
9.3.2 System model based on a radial basis function
neural net 187
9.4 Simulation results 192
9.4.1 MPC using SQP 193
9.4.2 MPC using interval search 194
9.5 Real-time implementation using a transputer system 195
9.5.1 Hardware structure 195
9.5.2 Software structure and parallel processing 198
9.6 Results 199
Contents ix
9.7 Conclusions 201
9.8 References 202
10 Fuzzy-neural control in intensive-care blood pressure
management DA. Linkens 203
10.1 Introduction 203
10.2 Fuzzy logic based reasoning 205
10.3 Functional fuzzy-neural control 205
10.3.1 Neural network based fuzzy control 206
10.3.2 Hybrid neural network fuzzy controller 208
10.4 Structural fuzzy-neural control 211
10.4.1 Simplified fuzzy control algorithm 211
10.4.2 Counter propagation network (CPN) fuzzy
controller 213
10.4.3 Radial basis function (RBF) fuzzy control 215
10.4.4 Cerebellar model articulation controller (CMAC)
fuzzy control 218
10.5 Multivariable modelling of the cardiovascular system 220
10.6 Experimental results 223
10.7 Conclusions 226
10.8 References 227
11 Neural networks and system identification 229
S.A. Billings and S. Chen
11.1 Introduction 229
11.2 Problem formulation 230
11.3 Learning algorithms for multi-layered neural networks 231
11.3.1 The multi-layered perceptron 231
11.3.2 Backpropagation 233
11.3.3 Prediction error learning algorithms 234
11.4 Radial basis function networks 236
11.4.1 Learning algorithms for radial basis function
networks 237
11.4.2 Backpropagation vs RPE vs the hybrid
clustering algorithm 238
11.5 The functional link or extended model set network 239
11.6 Properties of neural networks 240
11.6.1 Network expansions 240
11.6.2 Model validation 240
11.6.3 Noise and bias 242
11.6.4 Network node assignment 244
11.6.5 Network complexity 245
11.6.6 Metrics of network performance 246
11.7 System identification 247
11.7.1 Neural network models 247
11.7.2 The NARMAX methodology 247
11.8 Conclusions 248
11.9 Acknowledgments 249