A Study On Power Transformer Comprehensive Assessment And Power Transformer Decision Method On Condition- Based Maintenance Strategy

Since power transformer is an important kind of electric power equipment in power system, research of fuzzy comprehensive assessment (FCA) for its operating state and method of decision-making strategy with condition-based maintenance strategy (CBMS) used in the power transformer has great significance to improve the safety and reliability of power transformer and to keep the whole power system safe and reliable.This dissertation mainly discusses about fuzzy comprehensive assessment about operating state and the decision-making method using condition-based maintenance strategy for oil-immersed power transformer. Due to the characteristics of small sample data and lack of information during the fault diagnosis and operating state prediction of power transformer, this dissertation does many studies on the operating state comprehensive assessment and condition-based maintenance policy decision method for power transformer mainly using the fuzzy theory of numerical processing data. The main contents include:(1) Research on fuzzy comprehensive assessment method.Based on the research of state eigenvector and fault eigenvector of power transformer, this dissertation proposes the power transformer operating status of the technical scheme of comprehensive assessment based on the theory of fuzzy comprehensive assessment, building the model of state comprehensive assessment, establishing the method of state assessment matrix for power transformer using fuzzy AHP method, and some pieces of advice on how to improve it. It is presented that discussing about the application area and result analysis of the evaluation criterion of fuzzification used in this solution.(2) The research of the FCA algorithm of power transformer operating state.Power transformer operating state evaluation method has been built. By comparative analysis, the dissertation proposes the improved fuzzy comprehensive assessment method based on logarithmic function and trapezoidal fuzzy number. With prior knowledge and experiment analysis, the dissertation studies the feasibility of transformer condition assessment analysis of judgment matrix in the numerical with logarithmic function respectively and trapezoidal fuzzy number according to the establishment of the state equation method and evaluation system, proposing a new processing method of numerical fuzzification.(3) The study of the comprehensive assessment method based on fault diagnosis of power transformer.Considering the characteristics of the power transformer oil dissolved gas, this dissertation studies power transformer fault classification based on support vector machine and designs the fault classification model of support vector machine (SVM) under the decision tree, using vector projection method. By comparing the size of detectable level, the faults can be classified. The method, used in the process of classification of power transformer fault properties and fault parts, is effective and accurate. And it presents a comprehensive evaluation method based on the fault diagnosis of power transformer operating status.(4) The research of decision-making algorithm using CBMS for power transformer.On the basis of comprehensive evaluation for power transformer operating state, considering the nonlinear, multiple attribute and hybrid characteristics of power transformer maintenance policy decision criterion, this dissertation explores the transformer maintenance strategy decision scheme under the fuzzy comprehensive evaluation criterion system, and proposes the ideal solutions to the maintenance strategy based on hybrid multi-attribute criterion. The decision-making model of fuzzy TOPSIS has been established by using the precise number, interval number and fuzzy number, type and cost and efficiency indicators of coexisting criterion system. As a result, the effectiveness of decision and practicability are improved.This dissertation systematically studies the fuzzy comprehensive evaluation algorithm for operating state and decision-making method strategy using condition-based maintenance strategy for the power transformer. It has great theoretical significance and practical value for its development.