Kocijan Modelling and Control of Dynamic Systems Using Gaussian Process Models

1. Auflage 2016
ISBN: 978-3-319-21021-6
Verlag: Springer Nature Switzerland
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 281 Seiten

Reihe: Advances in Industrial Control

This monograph opens up new horizons for engineers and researchers in academia and in industry dealing with or interested in new developments in the field of system identification and control. It emphasizes guidelines for working solutions and practical advice for their implementation rather than the theoretical background of Gaussian process (GP) models. The book demonstrates the potential of this recent development in probabilistic machine-learning methods and gives the reader an intuitive understanding of the topic. The current state of the art is treated along with possible future directions for research.Systems control design relies on mathematical models and these may be developed from measurement data. This process of system identification, when based on GP models, can play an integral part of control design in data-based control and its description as such is an essential aspect of the text. The background of GP regression is introduced first with system identification and incorporation of prior knowledge then leading into full-blown control. The book is illustrated by extensive use of examples, line drawings, and graphical presentation of computer-simulation results and plant measurements. The research results presented are applied in real-life case studies drawn from successful applications including: a gas-liquid separator control; urban-traffic signal modelling and reconstruction; and prediction of atmospheric ozone concentration. A MATLAB® toolbox, for identification and simulation of dynamic GP models is provided for download.


Juš Kocijan is a senior research fellow at the Department of Systems and Control, Jozef Stefan Institute, the leading Slovenian research institute in the field of natural sciences and engineering, and a Professor of Electrical Engineering at the University of Nova Gorica, Slovenia. His past experience in the field of control engineering includes teaching and research at the University of Ljubljana and visiting research and teaching posts at several European universities and research institutes. He has been active in applied research in automatic control through numerous domestic and international research grants and projects, in a considerable number of which he acted as project leader. His research interests include the modelling of dynamic systems with Gaussian process models, control based on Gaussian process models, multiple-model approaches to modelling and control, applied nonlinear control, Individual Channel Analysis and Design. His other experience includes: serving as one of the editors of the Engineering Applications of Artificial Intelligence journal and on the editorial boards of other research journals, serving as a member of IFAC Technical committee on Computational Intelligence in Control, actively participating as a member of numerous scientific-meeting international programme and organising committees. Prof. Kocijan is a member of various national and international professional societies in the field of automatic control, modelling and simulation.

Autoren/Hrsg.

Kocijan, Jus

Kocijan, Juš

Inhaltsverzeichnis

1;Series Editors’ Foreword;6
2;Preface;8
3;Contents;11
4;Symbols and Notation;13
5;Acronyms;15
6;1 Introduction;17
6.1;1.1 Introduction to Gaussian-Process Regression;19
6.1.1;1.1.1 Preliminaries;19
6.1.2;1.1.2 Gaussian-Process Regression;23
6.2;1.2 Relevance;32
6.3;1.3 Outline of the Book;33
6.4;References;34
7;2 System Identification with GP Models;37
7.1;2.1 The Model Purpose;41
7.2;2.2 Obtaining Data---Design of the Experiment ƒ;42
7.3;2.3 Model Setup;44
7.3.1;2.3.1 Model Structure;44
7.3.2;2.3.2 Selection of Regressors;49
7.3.3;2.3.3 Covariance Functions;51
7.4;2.4 GP Model Selection;63
7.4.1;2.4.1 Bayesian Model Inference;64
7.4.2;2.4.2 Marginal Likelihood---Evidence Maximisation;66
7.4.3;2.4.3 Estimation and Model Structure;72
7.4.4;2.4.4 Selection of Mean Function;75
7.4.5;2.4.5 Asymptotic Properties of GP Models;77
7.5;2.5 Computational Implementation;78
7.5.1;2.5.1 Direct Implementation;78
7.5.2;2.5.2 Indirect Implementation;80
7.5.3;2.5.3 Evolving GP Models;86
7.6;2.6 Validation;91
7.7;2.7 Dynamic Model Simulation;96
7.7.1;2.7.1 Numerical Approximation;97
7.7.2;2.7.2 Analytical Approximation of Statistical Moments with a Taylor Expansion;97
7.7.3;2.7.3 Unscented Transformation;98
7.7.4;2.7.4 Analytical Approximation with Exact Matching of Statistical Moments;99
7.7.5;2.7.5 Propagation of Uncertainty;100
7.7.6;2.7.6 When to Use Uncertainty Propagation?;102
7.8;2.8 An Example of GP Model Identification;103
7.9;References;111
8;3 Incorporation of Prior Knowledge;119
8.1;3.1 Different Prior Knowledge and Its Incorporation;119
8.1.1;3.1.1 Changing Input--Output Data;120
8.1.2;3.1.2 Changing the Covariance Function;122
8.1.3;3.1.3 Combination with the Presumed Structure;122
8.2;3.2 Wiener and Hammerstein GP Models;123
8.2.1;3.2.1 GP Modelling Used in the Wiener Model;124
8.2.2;3.2.2 GP Modelling Used in the Hammerstein Model;129
8.3;3.3 Incorporation of Local Models;134
8.3.1;3.3.1 Local Models Incorporated into a GP Model;138
8.3.2;3.3.2 Fixed-Structure GP Model;148
8.4;References;159
9;4 Control with GP Models;163
9.1;4.1 Control with an Inverse Dynamics Model;166
9.2;4.2 Optimal Control;171
9.3;4.3 Model Predictive Control;174
9.4;4.4 Adaptive Control;202
9.5;4.5 Gain Scheduling;204
9.6;4.6 Model Identification Adaptive Control;209
9.7;4.7 Control Using Iterative Learning;214
9.8;References;219
10;5 Trends, Challenges and Research Opportunities;225
10.1;References;227
11;6 Case Studies;229
11.1;6.1 Gas--Liquid Separator Modelling and Control;230
11.2;6.2 Faulty Measurements Detection and Reconstruction in Urban Traffic;246
11.3;6.3 Prediction of Ozone Concentration in the Air;257
11.4;References;266
12;Appendix A Mathematical Preliminaries;269
13;Appendix B Predictions;273
14;Appendix C Matlab Code;278
15;Index;279