Study On Condition Maintenance Decision Making Method Of Power Transformer Based On Reliability And Risk Evaluation

As one of the most important and key equipments, a power transformer is the core of energy conversion and transmission in the power grid. For a long time, the judgment of health state and operation condition of power transformer is carried out by preventive test and periodic maintenance. The time-based maintenance mechanism will lead to"over maintenance"or"owing maintenance"for lack of consideration of power transformer's insulation condition. Therefore, with an urgent request for reliability and economy of the power system, there is an inevitable trend that the condition based maintenance (CBM) will replace the periodic maintenance for the power transformer maintenance. On the basis of the failure mode and effect analysis (FMEA) of power transformer, the key problems, such as the reliability, the failure probability, the failure risk, and the decision making method of CBM, were studied in this thesis.①FMEA of the power transformer. After an in-depth and detailed analysis on structures and functions of the power transformer, the method combining hardware and functions was employed to divide the composition of the power transformer and to analyze all failure modes and effects, which avoided the complexity of FMEA only according to the system structure and laid the foundation for evaluating transformer reliability and failure risk.②Comprehensive evaluation method for the power transformer failure probability based on multi-characteristic parameters. By means of generalizing, systemizing power transformer failure modes and analyzing features of every characteristic parameter, this paper proposed the power transformer failure probability assessment methods based on fuzzy membership function and relative inferiority degree. Besides, in order to take full advantage of characteristic parameters of failure mode which features multi-characteristic parameters, the paper put forward a comprehensive evaluation method for the transformer failure probability based on improved analytic hierarchy process (IAHP). This method realized the organic integration of different assessment methods, and avoided the shortcoming of expert evaluation method to a certain extent, such as subjectivity in order that the assessment results better correspond to reality.③Reliability assessment method of the power transformer based on the fault tree analysis (FTA). To date, to assess the reliability of power transformers primarily depends on statistical data and results of the assessment do not accurately reflect the reliability of a transformer. Therefore, this thesis built a complete power transformer fault tree based on FTA and FMEA tables, and then proposed the reliability assessment method of the power transformer based on FTA method and with the importance degree of failure mode taken into account. The fuzzy analytic hierarchy process (FAHP) reduced the subjectivity in defining importance weight of failure mode.④Risk assessment method of the power transformer based on fuzzy multiple attribute decision making (FMADM). Grounded on an in-depth study of MADM theory and methods, the failure risk assessment index system of the power transformer was established, the triangular fuzzy number was employed to reduce assessment index subjectivity, and the risk assessment method of the power transformer based on FMADM was presented, which overcame the shortcomings of traditional methods, such as the influence exerted by artificial factors, difficulty in distinguishing the importance of the various indexes and no possibility in evaluating the risk grade owing to the same risk priority number (RPN).⑤CBM decision making and optimization method based on the grey situation decision making (GSDM). On the basis of an in-depth study of GSDM method and in combination with results of the reliability and the failure risk assessment of the power transformer, the CBM decision making method of the power transformer was proposed in this thesis, which overcame the defect of maintenance program based on a single index. To bridge the gap that maintenance programs cannot give the optimal maintenance time, a maintenance time optimization method based on Monte Carlo simulation (MCS) was proposed in this dissertation.