Available:*
Library | Item Barcode | Call Number | Material Type | Item Category 1 | Status |
---|---|---|---|---|---|
Searching... | 30000010124465 | TP155.75 C56 2007 | Open Access Book | Book | Searching... |
On Order
Summary
Summary
The latest advances in process monitoring, data analysis, and control systems are increasingly useful for maintaining the safety, flexibility, and environmental compliance of industrial manufacturing operations.
Focusing on continuous, multivariate processes, Chemical Process Performance Evaluation introduces statistical methods and modeling techniques for process monitoring, performance evaluation, and fault diagnosis.
This book introduces practical multivariate statistical methods and empirical modeling development techniques, such as principal components regression, partial least squares regression, input-output modeling, state-space modeling, and modeling process signals for trend analysis. Then the authors examine fault diagnosis techniques based on episodes, hidden Markov models, contribution plots, discriminant analysis, and support vector machines. They address controller process evaluation and sensor failure detection, including methods for differentiating between sensor failures and process upset. The book concludes with an extensive discussion on the use of data analysis techniques for the special case of web and sheet processes. Case studies illustrate the implementation of methods presented throughout the book.
Emphasizing the balance between practice and theory, Chemical Process Performance Evaluation is an excellent tool for comparing alternative techniques for process monitoring, signal modeling, and process diagnosis. The unique integration of process and controller monitoring and fault diagnosis facilitates the practical implementation of unified and automated monitoring and diagnosis technologies.
Table of Contents
Nomenclature | |
1 Introduction | p. 1 |
1.1 Motivation and Historical Perspective | p. 2 |
1.2 Outline | p. 4 |
2 Univariate Statistical Monitoring Techniques | p. 7 |
2.1 Statistics Concepts | p. 8 |
2.2 Univariate SPM Techniques | p. 11 |
2.2.1 Shewhart Control Charts | p. 11 |
2.2.2 Cumulative Sum (CUSUM) Charts | p. 18 |
2.2.3 Moving Average Monitoring Charts for Individual Measurements | p. 19 |
2.2.4 Exponentially Weighted Moving Average Chart | p. 22 |
2.3 Monitoring Tools for Autocorreleated Data | p. 22 |
2.3.1 Monitoring with Charts of Residuals | p. 26 |
2.3.2 Monitoring with Detecting Changes in Model Parameters | p. 27 |
2.4 Limitations of Univariate Monitoring Techniques | p. 32 |
2.5 Summary | p. 35 |
3 Multivariate Statistical Monitoring Techniques | p. 37 |
3.1 Principal Components Analysis | p. 37 |
3.2 Canonical Variates Analysis | p. 43 |
3.3 Independent Component Analysis | p. 43 |
3.4 Contribution Plots | p. 46 |
3.5 Linear Methods for Diagnosis | p. 48 |
3.5.1 Clustering | p. 48 |
3.5.2 Discriminant Analysis | p. 50 |
3.5.3 Fisher's Discriminant Analysis | p. 53 |
3.6 Nonlinear Methods for Diagnosis | p. 58 |
3.6.1 Neural Networks | p. 58 |
3.6.2 Kernel-Based Techniques | p. 64 |
3.6.3 Support Vector Machines | p. 66 |
3.7 Summary | p. 69 |
4 Empirical Model Development | p. 73 |
4.1 Regression Models | p. 75 |
4.2 PCA Models | p. 78 |
4.3 PLS Regression Models | p. 79 |
4.4 Input-Output Models of Dynamic Processes | p. 83 |
4.5 State-Space Models | p. 89 |
4.6 Summary | p. 97 |
5 Monitoring of Multivariate Processes | p. 99 |
5.1 SPM Methods Based on PCA | p. 100 |
5.2 SPM Methods Based on PLS | p. 105 |
5.3 SPM Using Dynamic Process Models | p. 108 |
5.4 Other MSPM Techniques | p. 112 |
5.5 Summary | p. 114 |
6 Characterization of Process Signals | p. 115 |
6.1 Wavelets | p. 115 |
6.1.1 Fourier Transform | p. 116 |
6.1.2 Continuous Wavelet Transform | p. 119 |
6.1.3 Discrete Wavelet Transform | p. 123 |
6.2 Filtering and Outlier Detection | p. 127 |
6.2.1 Simple Filters | p. 128 |
6.2.2 Wavelet Filters | p. 131 |
6.2.3 Robust Filter | p. 133 |
6.3 Signal Representation by Fuzzy Triangular Episodes | p. 135 |
6.4 Development of Markovian Models | p. 138 |
6.4.1 Markov Chains | p. 139 |
6.4.2 Hidden Markov Models | p. 141 |
6.5 Wavelet-Domain Hidden Markov Models | p. 145 |
6.6 Summary | p. 147 |
7 Process Fault Diagnosis | p. 149 |
7.1 Fault Diagnosis Using Triangular Episodes and HMMs | p. 149 |
7.1.1 CSTR Simulation | p. 152 |
7.1.2 Vacuum Column | p. 155 |
7.2 Fault Diagnosis Using Wavelet-Domain HMMs | p. 157 |
7.2.1 pH Neutralization Simulation | p. 161 |
7.2.2 CSTR Simulation | p. 164 |
7.3 Fault Diagnosis Using HMMs | p. 166 |
7.3.1 Case Study of HTST Pasteurization Process | p. 167 |
7.4 Fault Diagnosis Using Contribution Plots | p. 174 |
7.5 Fault Diagnosis with Statistical Methods | p. 179 |
7.6 Fault Diagnosis Using SVM | p. 191 |
7.7 Fault Diagnosis with Robust Techniques | p. 192 |
7.7.1 Robust Monitoring Strategy | p. 192 |
7.7.2 Pilot-Scale Distillation Column | p. 198 |
7.8 Summary | p. 202 |
8 Sensor Failure Detection and Diagnosis | p. 203 |
8.1 Sensor FDD Using PLS and CVSS Models | p. 204 |
8.2 Real-Time Sensor FDD Using PCA-Based Techniques | p. 215 |
8.2.1 Methodology | p. 218 |
8.2.2 Case Study | p. 224 |
8.3 Summary | p. 230 |
9 Controller Performance Monitoring | p. 231 |
9.1 Single-Loop Controller Performance Monitoring | p. 233 |
9.2 Multivariable Controller Performance Monitoring | p. 237 |
9.3 CPM for MPC | p. 238 |
9.4 Summary | p. 248 |
10 Web and Sheet Processes | p. 251 |
10.1 Traditional Data Analysis | p. 252 |
10.1.1 MD/CD Decomposition | p. 252 |
10.1.2 Time Dependent Structure of Profile Data | p. 256 |
10.2 Orthogonal Decomposition of Profile Data | p. 257 |
10.2.1 Gram Polynomials | p. 259 |
10.2.2 Principal Components Analysis | p. 262 |
10.2.3 Flatness of Scanner Data | p. 264 |
10.3 Controller Performance | p. 268 |
10.3.1 MD Control Performance | p. 269 |
10.3.2 Model-Based CD Control Performance | p. 271 |
10.4 Summary | p. 274 |
Bibliography | p. 277 |
Index | p. 305 |