Condition monitoring and assessment of power transformers using computational intelligence / W.H. Tang, Q.H. Wu.

Includes bibliographical references and index.

1. Introduction -- 2. Evolutionary Computation -- 3. Methodologies Dealing With Uncertainty -- 4. Thermoelectric Analogy Thermal Models of Power Transformers -- 5. Thermal Model Parameter Identification and Verification Using Genetic Algorithm -- 6. Transformer Condition Assessment Using Dissolved Gas Analysis -- 7. Fault Classification for Dissolved Gas Analysis Using Genetic Programming -- 8. Dealing with Uncertainty for Dissolved Gas Analysis -- 9. Winding Frequency Response Analysis for Power Transformers -- 10. Winding Parameter Identification Using an Improved Particle Swarm Optimiser -- 11. Evidence-Based Winding Condition Assessment -- --

1. Introduction -- 1.1. Background -- 1.2. Main Aspects of Transformer Condition Monitoring and Assessment -- 1.2.1. Thermal Modelling -- 1.2.2. Dissolved Gas Analysis -- 1.2.3. Frequency Response Analysis -- 1.2.4. Partial Discharge Analysis -- 1.3. Drawbacks of Conventional Techniques -- 1.3.1. Inaccuracy of Empirical Thermal Models -- 1.3.2. Uncertainty in Dissolved Gas Analysis -- 1.3.3. Intricate Issues in Winding Deformation Diagnosis -- 1.4. Modelling Transformer and Processing Uncertainty Using Computational Intelligence -- 1.5. Contents of this Book -- 1.6. Summary -- 2. Evolutionary Computation -- 2.1. The Evolutionary Algorithms of Computational Intelligence -- 2.1.1. Objectives of Optimisation -- 2.1.2. Overview of Evolutionary Computation -- 2.2. Genetic Algorithm -- 2.2.1. Principles of Genetic Algorithms -- 2.2.2. Main Procedures of a Simple Genetic Algorithm -- 2.2.3. Implementation of a Simple Genetic Algorithm -- 2.3. Genetic Programming -- 2.3.1. Background of Genetic Programming -- 2.3.2. Implementation Processes of Genetic Programming -- 2.4. Particle Swarm Optimisation -- 2.4.1. Standard Particle Swarm Optimisation -- 2.4.2. Particle Swarm Optimisation with Passive Congregation -- 2.5. Summary -- 3. Methodologies Dealing With Uncertainty -- 3.1. The Logical Approach of Computational Intelligence -- 3.2. Evidential Reasoning -- 3.2.1. The Original Evidential Reasoning Algorithm -- 3.2.2. The Revised Evidential Reasoning Algorithm -- 3.3. Fuzzy Logic -- 3.3.1. Foundation of Fuzzy Logic -- 3.3.2. An Example of a Fuzzy Logic System -- 3.4. Bayesian Networks -- 3.4.1. The Bayes' Theorem -- 3.4.2. Bayesian Networks -- 3.4.3. Parameter Learning to Form a Bayesian Network -- 3.5. Summary -- 4. Thermoelectric Analogy Thermal Models of Power Transformers -- 4.1. Introduction -- 4.2. Conventional Thermal Models in IEC and IEEE Regulations -- 4.2.1. Steady-State Temperature Models -- 4.2.2. Transient-State Temperature Models -- 4.2.3. Hot-Spot Temperature Rise in Steady State -- 4.2.4. Hot-Spot Temperature Rise in Transient State -- 4.3. The Thermoelectric Analogy Theory -- 4.4. A Comprehensive Thermoelectric Analogy Thermal Model -- 4.4.1. Heat Transfer Schematics of Transformers -- 4.4.2. Derivation of a Comprehensive Heat Equivalent Circuit -- 4.5. Parameter Estimation of a Thermoelectric Analogy Model -- 4.5.1. Heat Generation Process -- 4.5.2. Heat Transfer Parameter -- 4.5.3. Operation Scheme of Winding Temperature Indicator -- 4.5.4. Time Constant Variation in a Heat Transfer Process -- 4.6. Identification of Thermal Model Parameters -- 4.7. A Simplified Thermoelectric Analogy Thermal Model -- 4.7.1. Derivation of a Simplified Heat Equivalent Circuit -- 4.7.2. Hot-Spot Temperature Calculation -- 4.8. Summary -- 5. Thermal Model Parameter Identification and Verification Using Genetic Algorithm -- 5.1. Introduction -- 5.2. Unit Conversion for Heat Equivalent Circuit Parameters -- 5.3. Fitness Function for Genetic Algorithm Optimisation -- 5.4. Parameter Identification and Verification for the Comprehensive Thermal Model -- 5.4.1. Estimation of Heat Transfer Parameters -- 5.4.2. Parameter Identification Using Genetic Algorithm -- 5.4.3. Verification of Identified Thermal Parameters Against Factory Heat Run Tests -- 5.4.4. Comparison between Modelling Results of Artificial Neural Network and Genetic Algorithm -- 5.5. Parameter Identification and Verification for the Simplified Thermal Model -- 5.5.1. Identification of Parameters Using Genetic Algorithm -- 5.5.2. Verification of Derived Parameters with Rapidly Changing Loads -- 5.5.3. Simulations of Step Responses Compared with Factory Heat Run -- 5.5.4. Hot-Spot Temperature Calculation -- 5.5.5. Error Analysis -- 5.6. Summary -- 6. Transformer Condition Assessment Using Dissolved Gas Analysis -- 6.1. Introduction -- 6.2. Fundamental of Dissolved Gas Analysis -- 6.2.1. Gas Evolution in a Transformer -- 6.2.2. Key Gas Method -- 6.2.3. Determination of Combustible Gassing Rate -- 6.2.4. Gas Ratio Methods -- 6.2.5. Fault Detectability Using Dissolved Gas Analysis -- 6.3. Combined Criteria for Dissolved Gas Analysis -- 6.4. Intelligent Diagnostic Methods for Dissolve Gas Analysis -- 6.5. Summary -- 7. Fault Classification for Dissolved Gas Analysis Using Genetic Programming -- 7.1. Introduction -- 7.2. Bootstrap -- 7.3. The Cybernetic Techniques of Computational Intelligence -- 7.3.1. Artificial Neural Network -- 7.3.2. Support Vector Machine -- 7.3.3. K-Nearest Neighbour -- 7.4. Results and Discussions -- 7.4.1. Process DGA Data Using Bootstrap -- 7.4.2. Feature Extraction with Genetic Programming -- 7.4.3. Fault Classification Results and Comparisons -- 7.5. Summary -- 8. Dealing with Uncertainty for Dissolved Gas Analysis -- 8.1. Introduction -- 8.2. Dissolved Gas Analysis Using Evidential Reasoning -- 8.2.1. A Decision Tree Model under an Evidential Reasoning Framework -- 8.2.2. An Evaluation Analysis Model based upon Evidential Reasoning -- 8.2.3. Determination of Weights of Attributes and Factors -- 8.2.4. Evaluation Examples under an Evidential Reasoning Framework -- 8.3. A Hybrid Diagnostic Approach Combining Fuzzy Logic and Evidential Reasoning -- 8.3.1. Solution to Crispy Decision Boundaries -- 8.3.2. Implementation of the Hybrid Diagnostic Approach -- 8.3.3. Tests and Results -- 8.4. Probabilistic Inference Using Bayesian Networks -- 8.4.1. Knowledge Transformation into a Bayesian Network -- 8.4.2. Results and Discussions -- 8.5. Summary -- 9. Winding Frequency Response Analysis for Power Transformers -- 9.1. Introduction -- 9.2. Transformer Transfer Function -- 9.3. Frequency Response Analysis Methods -- 9.3.1. Low Voltage Impulse -- 9.3.2. Sweep Frequency Response Analysis -- 9.4. Winding Models Used for Frequency Response Analysis -- 9.5. Transformer Winding Deformation Diagnosis -- 9.5.1. Comparison Techniques -- 9.5.2. Interpretation of Frequency Response Measurements -- 9.6. Summary -- 10. Winding Parameter Identification Using an Improved Particle Swarm Optimiser -- 10.1. Introduction -- 10.2. A Ladder Network Model for Frequency Response Analysis -- 10.3. Model-Based Approach to Parameter Identification and its Verification -- 10.3.1. Derivation of Winding Frequency Responses -- 10.3.2. Fitness Function Used by PSOPC -- 10.4. Simulations and Discussions -- 10.4.1. Test Simulations of Frequency Response Analysis -- 10.4.2. Winding Parameter Identification -- 10.4.3. Results and Discussions -- 10.5. Summary -- 11. Evidence-Based Winding Condition Assessment -- 11.1. Knowledge Transformation with Revised Evidential Reasoning Algorithm -- 11.2. A Basic Evaluation Analysis Model -- 11.3. A General Evaluation Analysis Model -- 11.4. Results and Discussions -- 11.4.1. An Example Using the Basic Evaluation Analysis Model -- 11.4.2. Aggregation of Subjective Judgements Using the General Evaluation Analysis Model -- 11.5. Summary -- Appendix: A Testing to BS171 for Oil-Immersed Power Transformers.

"In recent years, rapid changes and improvements have been witnessed in the field of transformer condition monitoring and assessment, especially with the advances in computational intelligence techniques. Condition Monitoring and Assessment of Power Transformers Using Computational Intelligence applies a broad range of computational intelligence techniques to deal with practical transformer operation problems. The approaches introduced are presented in a concise and flowing manner, tackling complex transformer modelling problems and uncertainties occurring in transformer fault diagnosis. Condition Monitoring and Assessment of Power Transformers Using Computational Intelligence covers both the fundamental theories and the most up-to-date research in this rapidly changing field. Many examples have been included that use real-world measurements and realistic operating scenarios of power transformers to fully illustrate the use of computational intelligence techniques for a variety of transformer modelling and fault diagnosis problems. Condition Monitoring and Assessment of Power Transformers Using Computational Intelligence is a useful book for professional engineers and postgraduate students. It also provides a firm foundation for advanced undergraduate students in power engineering."--Publisher's website.

Machine converted from AACR2 source record.