Diagnosis of Process Nonlinearities and Valve Stiction (eBook)

Data Driven Approaches

Ali Ahammad Shoukat Choudhury, Sirish L. Shah, Nina F. Thornhill (Autoren)

eBook Download: PDF

2008 | 2008
XX, 286 Seiten
Springer Berlin (Verlag)
978-3-540-79224-6 (ISBN)

Infos

were published in the series as the contributed volume, Process Control Performance Assessment: From Theory to Implementation with Andrzej Ordys, Damian Uduehi, and Michael Johnson as Editors (ISBN 978-1-84628-623-0, 2007). Along with this good progress in process controller assessment methods, researchers have also been investigating techniques to diagnose what is causing the process or control loop degradation. This requires the use of on-line data to identify faults via new diagnostic indicators of typical process problems. A significant focus of some of this research has been the issue of valve problems; a research direction that has been motivated by some industrial statistics that show up to 40% of control loops having performance degradation attributable to valve problems. Shoukat Choudhury, Sirish Shah, and Nina Thornhill have been very active in this research field for a number of years and have written a coherent and consistent presentation of their many research results as this monograph, Diagnosis of Process Nonlinearities and Valve Stiction. The Advances in Industrial Control series is pleased to welcome this new and substantial contribution to the process diagnostic literature. The reader will find the exploitation of the extensive process data archives created by today's process computer systems one theme in the monograph. From another viewpoint, the use of higher-order statistics could be considered to provide a continuing link to the earlier methods of the statistical process control paradigm.

M. A. A. Shoukat Choudhury received his B. Sc. Engineering (Chemical) from Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh in 1996. He was awarded a gold medal for his outstanding results in B. Sc. Engineering. He obtained an M. Sc. Engineering (Chemical) in 1998 from the same university. He has completed his PhD degree in process control (Chemical Engineering) at the University of Alberta, Canada. For his outstanding research performance during the course of PhD program he has been awarded several awards such as University of Alberta PhD Dissertation Fellowship, Andrew Stewart Memorial Prize and ISA Educational Foundation Scholarship. He is the principal inventor of an internation patent (applied, 2005) on 'Methods for Detection and Quantification of Control Valve Stiction'. The methodologies and algorithms described in this patent are implemented and available in the commercial software ProcessDoctor from Matrikon Inc. His main research interests include diagnosis of poor control performance, stiction in control valves, data compression, control loop performance assessment and monitoring, and diagnosis of plant wide oscillations.

Sirish Shah received his B.Sc. degree in control engineering from Leeds University in 1971, a M.Sc. degree in automatic control from UMIST, Manchester in 1972, and a Ph.D. degree in process control (chemical engineering) from the University of Alberta in 1976. During 1977 he worked as a computer applications engineer at Esso Chemicals in Sarnia, Ontario. Since 1978 he has been with the University of Alberta, where currently holds the NSERC-Matrikon-ASRA Senior Industrial Research Chair in Computer Process Control. In 1989, he was the recipient of the Albright & Wilson Americas Award of the Canadian Society for Chemical Engineering in recognition of distinguished contributions to chemical engineering. He has held visiting appointments at Oxford University and Balliol College as a SERC fellow in 1985-86 and at Kumamoto University, Japan as a senior research fellow of the Japan Society for the Promotion of Science (JSPS) in 1994. The main area of his current research is process and performance monitoring, system identification and design and implementation of softsensors. He has recently co-authored a book titled, Performance Assessment of Control Loops: Theory and Applications. He has held consulting appointments with a wide variety of process Industries and has also taught many industrial courses.

Inhalt

Preface 7
Author Biographies 11
Contents 13
Introduction 21
1.1 Concepts in Data-Driven Analysis of Chemical Processes 22
1.1.1 Linear and Nonlinear Time Series 23
1.1.2 Statistics and Randomness 23
1.1.3 Frequency Content and Spectral Methods 26
1.2 Nonlinearity in Control Valves 28
1.3 The Layout of the Book 30
1.3.1 Part I Higher-Order Statistics 30
1.3.2 Part II Data Quality – Compression and Quantization 30
1.3.3 Part III Nonlinearity and Control Performance 31
1.3.4 Part IV Control Valve Stiction – Definition, Modelling, Detection and Quantification 32
1.3.5 Part V Plant-wide Oscillations – Detection and Diagnosis 33
1.3.6 References 34
1.4 Summary 34
Part I 36
Higher-Order Statistics: Preliminaries 37
2.1 Introduction 37
2.2 Time Domain Analysis 38
2.2.1 Moments 38
2.2.2 Cumulants 40
2.2.3 The Relationship Between Moments and Cumulants 42
2.2.4 Properties of Moments and Cumulants 42
2.2.5 Moments and Cumulants of Stationary Signals 45
2.3 Spectral Analysis 45
2.3.1 Power Spectrum, 46
2.3.2 Bispectrum, 47
2.4 Summary 48
Bispectrum and Bicoherence 49
3.1 Bispectrum 49
3.1.1 Estimation of the Bispectrum 50
3.1.2 Properties of Estimators and Asymptotic Behaviour 52
3.1.3 Bicoherence or Normalized Bispectrum 54
3.1.4 Properties of Bispectrum and Bicoherence 55
3.2 Bispectrum or Bicoherence Estimation Issues 57
3.2.1 Choice of Window Function 58
3.2.2 Choice of Data Length, Segment Length and Fourier Transform Length 60
3.3 Summary 61
Part II 63
Impact of Data Compression and Quantization on Data- Driven Process Analyses 65
4.1 Introduction 65
4.2 Data Compression Methods 67
4.2.1 Overview of Data Compression 67
4.2.2 Box-Car (BC) Algorithm 67
4.2.3 Backward-Slope (BS) Algorithm 67
4.2.4 Combined Box-Car and Backward-Slope (BCBS) Method 69
4.2.5 Swinging Door Compression Algorithm 69
4.2.6 The Compression Factor 69
4.3 Measures of Data Quality 70
4.3.1 Statistical Measures 70
4.3.2 Nonlinearity Measures 71
4.3.3 Performance Index (Harris) Measures 71
4.4 Process Data for Compression Comparison 72
4.4.1 Industrial Example 1 72
4.4.2 Industrial Example 2 75
4.5 Results and Discussions for Industrial Example 2 76
4.5.1 Visual Observations 76
4.5.2 Statistical Properties 77
4.5.3 Nonlinearity Assessment 78
4.5.4 Performance (Harris) Index 78
4.6 Summary of Data Quality Measures 79
4.7 Automated Detection of Compression 79
4.7.1 Motivation 79
4.7.2 Compression Detection Procedure 80
4.7.3 Implementation Considerations 81
4.8 A Recommendation for Harmless Storing of Data 83
4.9 Quantization 83
4.10 Summary 85
Part III 87
Measures of Nonlinearity – A Review 89
5.1 Definition of Nonlinear Systems 89
5.2 Nonlinearity in Process Time Trends 90
5.3 Various Measures of Nonlinearity 90
5.3.1 Model-Based Measures of Nonlinearity 91
5.3.2 Time Series-Based Measures of Nonlinearity 91
5.4 Summary 95
Linear or Nonlinear? A Bicoherence-Based Measure of Nonlinearity 97
6.1 Introduction 97
6.2 Bispectrum and Bicoherence 98
6.2.1 Spurious Peaks in 98
the Estimated Bicoherence 98
6.2.2 Illustrative Example 1 99
6.2.3 How to Choose 100
6.3 Test of Gaussianity and Linearity of a Signal 101
6.3.1 Total Nonlinearity Index (TNLI) 105
6.4 Illustrative Example 2: Bicoherence of a Linear and a Nonlinear Signal 105
6.5 Illustrative Example 3: Bicoherence of a Nonlinear Sinusoid Signal with Noise 107
6.5.1 Mild Nonlinearity ( nl = 108
0.05) 108
6.5.2 Strong Nonlinearity ( nl = 109
0.25) 109
6.5.3 Extent of Nonlinearity and Effect of Noise 110
6.6 Summary 111
A Nonlinearity Measure Based on Surrogate Data Analysis 113
7.1 Surrogate Time Series 113
7.1.1 Nonlinearity Detection Using Surrogates 113
7.1.2 Predictability in Nonlinear Time Series 113
7.2 Algorithm for Nonlinearity Diagnosis 115
7.2.1 Construction of the Data Matrix for Nonlinear Prediction 115
7.2.2 Calculation of Prediction Error 116
7.2.3 Calculation of Surrogate Data 116
7.2.4 Statistical Testing 118
7.2.5 Algorithm Summary 118
7.3 Selection of the Parameter Values 119
7.3.1 Recommended Default Parameter Values 119
7.3.2 Choice of Embedding Parameters E and H 119
7.3.3 Choice of Parameters C and k 120
7.3.4 Default Data Ensemble Size, Q and Number of Samples Per Feature, S 121
7.3.5 Choice of the Number of Surrogates, M 121
7.4 Data-Preprocessing and End-Matching 122
7.4.1 False-Positive Results with Cyclic Data 122
7.4.2 End-Matching 123
7.4.3 Summary of the Data-Preprocessing Steps 124
7.4.4 Application to Oscillating Time Trends 124
7.5 Worked Examples 126
7.5.1 Identification of Nonlinear Root Causes 126
7.5.2 Application to the SE Asia Data Set 126
7.5.3 The Mechanism of Propagation in the SE Asia Process 126
7.5.4 An Example with No Nonlinearity 128
7.6 Summary 130
Nonlinearities in Control Loops 131
8.1 Process Nonlinearity 131
8.1.1 Nonlinearity of a Spherical Tank 131
8.1.2 Nonlinearities of a Continuous Stirred Tank Reactor (CSTR) 135
8.2 Nonlinear Valve Characteristic 137
8.2.1 Linear Valves 138
8.2.2 Equal Percentage Valves 138
8.2.3 Square-Root Valve 139
8.2.4 Remarks on Nonlinear Valve Characteristic 140
8.3 Nonlinear Disturbances 141
8.4 Summary 141
Diagnosis of Poor Control Performance 143
9.1 Introduction 143
9.2 Problem Description 144
9.3 Usual Causes of Poor Performance 145
9.4 Diagnosis of Poor Control Performance 146
9.4.1 Well Tuned Controller 147
9.4.2 Tightly Tuned Controller or Excessive Integral Action 148
9.4.3 Presence of an External Oscillatory Disturbance 149
9.4.4 Presence of Stiction 149
9.5 Industrial Case Studies 149
9.5.1 Stiction in a Furnace Dryer Temperature Control Valve 150
9.5.2 Valve Saturation 151
9.5.3 Valve Problems in Some Flow Control Loops 152
9.6 Summary 154
Part IV 156
Different Types of Faults in Control Valves 157
10.1 What Is a Control Valve 157
10.2 Faults in Control Valve 158
10.2.1 Oversized Valve 159
10.2.2 Undersized Valve 159
10.2.3 Corroded Valve Seat 159
10.2.4 Faulty Diaphragm 159
10.2.5 Packing Leakage 159
10.2.6 Valve Hysteresis 160
10.2.7 Valve Stiction 160
10.2.8 Large Deadband 160
10.2.9 Valve Saturation 161
10.3 Summary 161
Stiction: Definition and Discussions 163
11.1 Introduction 163
11.2 What Is Stiction? 163
11.2.1 Definition of Terms Relating to Valve Nonlinearity 164
11.2.2 Discussion of the Term ‘Stiction’ 165
11.2.3 A Formal Definition of Stiction 166
11.3 Practical Examples of Valve Stiction 168
11.4 Summary 171
Physics-Based Model of Control Valve Stiction 173
12.1 Introduction 173
12.2 Physical Modelling of Valve Friction 173
12.2.1 Physics of a Control Valve 173
12.2.2 Friction Model 174
12.2.3 Model Parameters 175
12.2.4 Detection of Zero Velocity 176
12.2.5 Model of the Pressure Chamber 176
12.3 Valve Simulation 177
12.3.1 Open-Loop Response 177
12.3.2 Closed-Loop Response 178
12.4 Summary 180
Data-Driven Model of Valve Stiction 181
13.1 One-Parameter Data-Driven Stiction Model 181
13.2 Two-Parameter Data-Driven Model of Valve Stiction 183
13.2.1 Model Formulation 183
13.2.2 Dealing with Stochastic or Noisy Control Signals 186
13.2.3 Open-Loop Response of the Model Under a Sinusoidal Input 186
13.2.4 Stiction in Reality 187
13.2.5 Closed-Loop Behaviour of the Model 187
13.3 Comparison of Physics-Based Model and Data- Driven Model 191
13.4 Summary 191
Describing Function Analysis 193
14.1 Introduction 193
14.2 Describing Function Analysis for Two-Parameter Stiction Model 194
14.2.1 Derivation of the Describing Function 194
14.3 Asymptotes of the Describing Function 197
14.4 Insights Gained from the Describing Function 198
14.4.1 The Impact of the Controller on the Limit Cycle 199
14.5 Summary 200
Automatic Detection and Quantification of Valve Stiction 201
15.1 Introduction 201
15.2 Stiction Detection – A Literature Review 202
15.3 Detection of Stiction Using Nonlinearity Information 203
and the 203
pv 203
– 203
op 203
Mapping 203
15.3.1 Detection of Loop Nonlinearity 204
15.3.2 Use of pv–op Plot 205
15.4 Stiction Quantification 207
15.4.1 Clustering Techniques of Quantifying Stiction 207
15.4.2 Fitted Ellipse Technique for Quantifying Stiction 210
15.5 An Illustrative Example 212
15.5.1 Validation of the Results 213
15.6 Automation of the Method 213
15.7 Simulation Results 215
15.7.1 A Worked Example 215
15.7.2 Distinguishing Limit Cycles Caused by Stiction and Those Caused by a Sinusoidal Disturbance 216
15.7.3 Detecting Stiction When Its Impact Propagates as Disturbance 218
15.8 Practical Implementation Issues 221
15.8.1 Bicoherence Estimation 221
15.8.2 Nonstationarity of the Data 221
15.8.3 Problems of Outliers and Abrupt Changes 221
15.8.4 Dealing with Short Length Data 222
15.8.5 Dealing with Longer Oscillations 222
15.8.6 Valve Nonlinearity 222
15.8.7 Filtering of the Data 223
15.8.8 Segmenting Data for pv–op Plot 224
15.9 Summary 224
Industrial Applications of the Stiction Quantification Algorithm 225
16.1 Industrial Case Studies 225
16.1.1 Loop 1: A Level Loop 225
16.1.2 Loop 2: A Linear-Level Control Loop 227
16.1.3 Loop 3: A Flow Control Loop 228
16.1.4 Loop 4: Flow Control Loop Cascaded with Level Control 229
16.1.5 Loop 5: A Pressure Control Loop 230
16.1.6 Loop 6: A Composition Control Loop 230
16.1.7 Loop 7: A Cascaded Flow Control Loop 231
16.1.8 Loop 8: A Temperature Control Loop 232
16.1.9 Loops 9 and 10 232
16.2 Online Compensation for Stiction 233
16.3 Summary 235
Confirming Valve Stiction 237
17.1 Methods to Confirm Valve Stiction 237
17.2 Gain Change Method for Confirming Valve Stiction 238
17.2.1 Distinguishing Stiction from External Oscillatory Disturbance 238
17.3 Describing Function Analysis 242
17.3.1 Comparison of Describing Function Analysis (DFA) Results with Simulation Results 245
17.4 Industrial Example 245
17.5 Summary 246
Part V 248
Detection of Plantwide Oscillations 249
18.1 Introduction 249
18.2 What is an Oscillation? 250
18.2.1 Units of Frequency 250
18.2.2 Examples of Oscillatory Signals 250
18.3 Detection of Oscillation(s) in a Single Time Series 251
18.3.1 The Power Spectrum 251
18.3.2 H ¨ agglund’s IAE Method 251
18.3.3 Autocovariance (ACF) Based Method 252
18.3.4 Other Methods 257
18.4 What are Plant-wide Oscillations? 257
18.5 Classification of Plant-wide Oscillations or Disturbances 257
18.5.1 Time scales 257
18.5.2 Oscillating and Non-oscillating Disturbances 258
18.6 Detection of Plant-wide Oscillations 258
18.6.1 High-Density Plots 258
18.6.2 ACF-Based Method 259
18.6.3 Power Spectral Correlation Map (PSCMAP) 259
18.6.4 Spectral Envelope Method 260
18.6.5 Spectral Decomposition Methods 261
18.7 Summary 270
Diagnosis of Plant-wide Oscillations 273
19.1 Root Cause Diagnosis of Plant-wide Oscillation 273
19.1.1 Finding a Nonlinear Root Cause of a Plant-Wide Disturbance 273
19.1.2 Finding a Linear Root Cause of a Plant-wide Disturbance 276
19.2 Industrial Case Study 1 – Eastman Chemical Plant 277
19.2.1 Data Description 278
19.2.2 Reduction of the Problem Size 278
19.2.3 Detection of Plant-wide Oscillation by PSCMAP 279
19.2.4 Nonlinearity Analysis Using Bicoherence-Based Indices 280
19.2.5 Diagnosis of the Problem in Loop LC2 282
19.3 Industrial Case Study 2 – SE Asia Refinery Data Analysis 283
19.3.1 Oscillation Detection by PSCMAP 284
19.3.2 Oscillation Detection by Spectral Envelope 285
19.3.3 Oscillation Diagnosis 286
19.4 Industrial Case Study 3 – Mitshubishi Chemical Corporation 286
19.4.1 Scope of the Analysis and Data Set 288
19.4.2 Oscillation-Detection Results 288
19.4.3 Oscillation Diagnosis 288
19.4.4 The Results of Maintenance on the PC1 and LI1 Loops 291
19.5 Summary 292
References 293
Copyright Acknowledgements 301
Index 303

Produkt-Details

• Erscheint lt. Verlag: 20.8.2008
• Reihe/Serie: Advances in Industrial Control
• Zusatzinfo: XX, 286 p. 313 illus., 115 illus. in color.
• Verlagsort: Berlin
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Mathematik ► Statistik
• Themenwelt: Technik ► Elektrotechnik / Energietechnik
• Themenwelt: Wirtschaft ► Betriebswirtschaft / Management
• Schlagworte: compression • Control • Diagnose • Higher Order Statistics • Linearity • Modeling • nonlinearity • quality • Quality Control, Reliability, Safety and Risk • Signal Processing • Stiction • Valves
• ISBN-10: 3-540-79224-4 / 3540792244
• ISBN-13: 978-3-540-79224-6 / 9783540792246