Bayesian networks in fault diagnosis: practice and applications

Contents......Page 8
Preface......Page 6
 1.1 Introduction......Page 18
 1.2 Overview of BNs......Page 20
 1.3 Procedures of Fault Diagnosis with BNs......Page 21
  1.3.1. BN structure modeling......Page 22
  1.3.2. BN parameter modeling......Page 25
  1.3.3. BN inference......Page 28
  1.3.4. Fault identification......Page 29
  1.3.5. Verification and validation......Page 31
  1.4.2. DBNs for fault diagnosis......Page 34
  1.4.3. OOBNs for fault diagnosis......Page 35
  1.4.4. Other BNs for fault diagnosis......Page 36
  1.5.1. Fault diagnosis for process systems......Page 37
  1.5.2. Fault diagnosis for energy systems......Page 39
  1.5.3. Fault diagnosis for structural systems......Page 40
  1.5.5. Fault diagnosis for network systems......Page 41
  1.6.1. Integrated big data and BN fault diagnosis methodology......Page 42
  1.6.3. Fast inference algorithms of BNs for online fault diagnosis......Page 43
  1.6.5. Fault identification rules......Page 44
 1.7 Conclusion......Page 45
 References......Page 46
2. Multi-Source Information Fusion-BasedFault Diagnosis of Ground-Source Heat Pump Using Bayesian Network......Page 58
 2.1 Introduction......Page 59
 2.2 Faults and Fault Symptoms......Page 61
   2.3.1.2. BN parameters......Page 64
   2.3.2.2. BN parameters......Page 67
  2.3.3. Multi-source information fusion-based fault diagnosis......Page 68
  2.4.1. Fault diagnosis using evidences from only sensor data......Page 71
  2.4.2. Fault diagnosis using evidences from sensor data and observed information......Page 73
 2.5. Conclusion......Page 76
 References......Page 77
 3.1. Introduction......Page 82
 3.2. System Description and Fault Analysis......Page 87
  3.3.1. Proposed fault diagnosis methodology......Page 91
  3.3.2. Signal feature extraction using FFT......Page 92
  3.3.3. Dimensionality reduction using PCA......Page 94
  3.3.4. Fault diagnosis using BNs......Page 96
  3.4.1. Simulation and experimental setup......Page 100
  3.4.2. Results......Page 102
 3.5. Conclusion......Page 108
 References......Page 109
 4.1. Introduction......Page 112
  4.2.1. Overview of OOBNs......Page 116
  4.2.2. Modeling methodology......Page 117
  4.2.3. Structure of OOBNs......Page 119
  4.2.4. Parameter of OOBNs......Page 121
  4.2.5. Model validation......Page 122
  4.2.6. Fault diagnosis and verification......Page 123
  4.3.1. Description of subsea production system......Page 124
  4.3.2. Fault diagnosis modeling......Page 126
  4.3.3. Results and discussion......Page 133
 4.4. Conclusion......Page 137
 References......Page 138
 5.1. Introduction......Page 142
 5.2. Faults Description......Page 145
 5.3. Modeling Methodology......Page 147
  5.3.1. DBNs structure modeling......Page 148
  5.3.2. DBN parameter modeling......Page 149
  5.4.1. Description of GMR control systems......Page 152
  5.4.2. Fault diagnosis modeling......Page 154
  5.4.3. Results and discussion......Page 157
 5.5. Conclusion......Page 162
 References......Page 163
6. An Integrated Safety Prognosis Model for Complex System Based on Dynamic Bayesian Network and Ant Colony Algorithm ......Page 168
 6.1. Introduction......Page 169
 6.2. Dynamic Bayesian Networks......Page 173
 6.3. Proposed Integrated Safety Prognosis Model......Page 174
  6.3.1. HAZOP model development......Page 175
  6.3.2. Degradation model development......Page 178
  6.3.3. DBN model development......Page 180
  6.3.4. Monitoring model development......Page 182
  6.3.5. Assessment model development......Page 183
  6.3.6. Risk evaluation model development......Page 184
  6.3.7. Prediction model development......Page 185
 6.4. Application to Gas Turbine Compressor System......Page 186
  6.5.1. The results of safety assessment......Page 192
  6.5.2. The results of risk evaluation......Page 195
  6.5.3. The results of safety prediction......Page 201
 6.6. Conclusion......Page 203
 References......Page 204
 Appendix......Page 207
 7.1. Introduction......Page 218
 7.2. MFM Modeling and Functional HAZOP Study......Page 222
  7.2.1. Traditional HAZOP study......Page 224
  7.2.2. Functional HAZOP study......Page 225
  7.2.3. Phase 1: MFM modeling of FCCU......Page 229
   7.2.3.2. Target decomposition of the reaction–regeneration unit......Page 230
   7.2.3.3. Analysis of the main components and functions of the regeneration–reaction......Page 232
  7.2.5. Phase 3: Functional HAZOP study results of FCCU......Page 233
  7.3.1. Dynamic Bayesian network......Page 237
  7.3.2. Integrated methodology procedure......Page 244
  7.4.1. Stage I: DBN modeling......Page 247
  7.4.2. Stage II: online fault diagnosis......Page 250
   7.4.3.2. DCS versus IFDS......Page 254
   7.4.3.3. Existing diagnosis methods versus IFDS......Page 256
 7.5. Conclusion......Page 258
 References......Page 259
8. DBN-Based Failure Prognosis Method Considering the Response of Protective Layers for Complex Industrial Systems......Page 262
 8.1. Introduction......Page 263
 8.2. DBN-Based Root Cause Analysis and Failure Prognosis Framework......Page 266
  8.3.1. Functional analysis of the protective layers......Page 270
  8.3.2. Extended DBN model for failure prognosis considering PL effect......Page 272
 8.4. DBN-Based Failure Prognosis Modeling for FGERS......Page 277
 8.5. Case Study......Page 285
  8.5.1. Failure prognosis in time dimension......Page 287
  8.5.2. Failure prognosis in space dimension......Page 290
 8.6. Conclusion......Page 292
 References......Page 293
 9.1. Introduction......Page 296
  9.2.1. BP-MLE method under incomplete data......Page 301
  9.2.2. BP-FS method under incomplete expert knowledge......Page 302
 9.3. Application of the Proposed Methods in an SAHP System......Page 304
  9.3.1. Structure of the BN......Page 305
  9.3.2. Parameter learning of conditional probabilities with incomplete data......Page 308
  9.3.3. Parameter estimation of prior probabilities with BP-FS method......Page 310
  9.4.1. Fault diagnosis using complete symptoms......Page 314
 9.5. Conclusion......Page 317
 References......Page 318
 10.1. Introduction......Page 322
 10.2. The Proposed Fault Diagnosis Methodology......Page 325
  10.3.1. Hydraulic control system of subsea blowout preventer......Page 326
   10.3.2.1. Develop Bayesian networks of operation procedures......Page 328
   10.3.2.2. Establish the Bayesian network of state decision nodes......Page 331
   10.3.2.3. Develop the entire Bayesian network......Page 334
  10.4.1. No faults in the closing process......Page 336
  10.4.4. One fault of locking system in blue pod......Page 338
 10.5. Conclusion......Page 342
 References......Page 343
11. A DBN-Based Risk Assessment Model for Prediction and Diagnosis of Offshore Drilling Incidents......Page 346
 11.1. Introduction......Page 347
 11.2. Manage Pressure Drilling Technology......Page 350
  11.3.1. Bayesian networks......Page 352
  11.3.2. Dynamic Bayesian networks......Page 353
 11.4. Development of a DBN-Based Risk Assessment Model......Page 354
  11.4.1. Step 1: Hazard identification......Page 356
   11.4.2.1. Mapping BT to BN......Page 358
   11.4.2.2. Simplified DBN model development......Page 359
   11.4.3.1. Predictive analysis......Page 363
   11.4.3.2. Diagnostic analysis......Page 364
  11.4.4. Validation of the model......Page 365
 11.5. Case Study......Page 371
  11.5.1. Risk identification for lost circulation......Page 374
  11.5.2. DBN modeling for the case......Page 375
   11.5.3.1. Risk evolution prediction......Page 381
   11.5.3.2. Root cause reasoning......Page 383
   11.5.3.3. Sensitivity analysis......Page 385
 11.6. Conclusions and Research Perspectives......Page 387
 References......Page 388
 12.1. Introduction......Page 392
  12.2.1. EEMD algorithm and feature extraction method......Page 395
  12.2.2. Bayesian network......Page 398
  12.3.1. The proposed methodology......Page 399
  12.3.2. Experiment and feature extraction......Page 400
  12.3.3. Bayesian network structure......Page 402
  12.3.4. Bayesian network parameters......Page 404
  12.4.1. Fault diagnosis only using fault features......Page 408
 12.5. Conclusion......Page 412
 References......Page 413
Index......Page 416