Plug-and-Play Monitoring and Performance Optimization for Industrial Automation Processes (eBook)

Hao Luo received the Ph.D. degree at the Institute for Automatic Control and Complex Systems (AKS) at the University of Duisburg-Essen, Germany, in 2016. His research interests include model-based and data-driven fault diagnosis, fault-tolerant systems and their industrial applications.

Acknowledgements 5
Contents 7
List of Figures 10
List of Tables 12
Nomenclature 13
1 Introduction 16
1.1 Background and motivation 16
1.1.1 FDI and FTC in complex industrial systems 17
1.1.2 PnP control concept 20
1.2 Objective of the work 21
1.3 Outline of the thesis 22
2 Basics of Process Monitoring Techniques 24
2.1 Mathematical description of automatic control processes 24
2.1.1 Description of nominal system behavior 24
2.1.2 Coprime factorization technique 25
2.1.3 Description of systems with disturbances 26
2.1.4 Description of systems with faults 26
2.2 Model-based residual generation techniques 27
2.2.1 Kernel representation and fault detection filter 27
2.2.2 Diagnostic observer 28
2.2.3 Parity space approach 29
2.2.4 Interconnections between DO and PS schemes 31
2.3 Data-driven residual generation techniques 32
2.3.1 SIM-aided process monitoring 32
2.3.2 Data-driven design of residual generator 33
2.4 Residual evaluation and decision making 35
2.4.1 Residual evaluation strategies 36
2.4.2 Threshold setting and decision making 37
2.5 Multivariate statistical process monitoring techniques 37
2.6 Concluding remarks 38
3 Basics of FTC Structure 39
3.1 Standard feedback control structure 39
3.2 Well-posedness and internal stability 40
3.2.1 Well-posedness 40
3.2.2 Internal stability 41
3.3 Image representation and state feedback control 43
3.4 Parameterization of stabilizing controllers 44
3.5 Model uncertainty and robustness 47
3.5.1 Small gain theorem 47
3.5.2 Coprime factor uncertainty 48
3.6 The fault-tolerant control architecture 51
3.7 Concluding remarks 53
4 PnP Process Monitoring and Control Architecture 54
4.1 Problem formulation 54
4.2 Scalability of feedback control systems 56
4.3 The PnP process monitoring and control architecture 59
4.3.1 The PnP-PMCA 59
4.3.2 Comparison with the fault-tolerant control architecture 61
4.3.3 Industrial implementation of the PnP-PMCA 63
4.4 PnP control strategies for new actuators and sensors 66
4.4.1 PnP control strategy for new actuators 66
4.4.2 PnP control strategy for new sensors 67
4.5 Concluding remarks 68
5 Real-Time Configuration Techniques for PnP Process Monitoring 69
5.1 Adaptive observer-based configuration 70
5.1.1 The canonical forms of LTI state-space systems 70
5.1.2 Adaptive configuration approach 72
5.2 Iterative configuration approach 79
5.2.1 The input/output normal form 81
5.2.2 Iterative configuration approach 84
5.3 Process monitoring with deterministic disturbance 91
5.3.1 Preliminaries related to the model-based solution 91
5.3.2 A data-driven process monitoring approach 93
5.4 Concluding remarks 95
6 Real-Time Configuration Techniques for PnP Performance Optimization 96
6.1 Control performance assessment system 96
6.2 Internal stability of the PnP-PMCA 99
6.2.1 Closed-loop dynamics of the PnP-PMCA 99
6.2.2 Constraints on closed-loop internal stability 101
6.3 Control performance optimization in PnP-PMCA 105
6.3.1 Iterative robustness optimization 105
6.3.2 Iterative tracking performance optimization 112
6.4 Convergence analysis 117
6.5 Concluding remarks 118
7 Benchmark Study and Real-Time Implementation 120
7.1 Application to rolling mill benchmark 120
7.1.1 General description of rolling mill system 120
7.1.2 PnP process monitoring and disturbance compensation system 124
7.1.3 Roll eccentricity monitoring and compensation module 125
7.1.4 Case study and simulation results 131
7.2 Real-time implementation on BLDC motor test rig 137
7.2.1 Description of the test rig 137
7.2.2 HIL simulation result 139
7.3 Concluding remarks 143
8 Conclusions and Future Work 145
A Proof of Theorem 4.2 147
Bibliography 150