Research On The Rimer Fault Diagnosis And Topsis State Forecasting Of Traction Transformer Insulation

With the rapid development of china's electrified railway, traction power supply system needs to become more secure and reliable. As the core of electric energy conversion and transmission in the electrified railway, the traction transformer is crucial to the traction power supply system. The running state of the traction transformer can directly affect the safety level of rail transport. Therefore it has a very important theoretical and practical significance to study the insulation fault diagnosis, state assessment and maintenance techniques for traction transformer. The state maintenance mechanism of the railway power supply equipment maintenance has not yet been completed by now. Transformer insulation fault diagnosis, state assessment and prediction are the theoretical basis of the state maintenance. The transformer insulation fault diagnosis method, state assessment method, state prediction method and traction transformer state maintenance were studied in the dissertation.A fault diagnosis method based on RIMER (belief rule-base inference methodology using the evidential reasoning approach) expert system is proposed to deal with the complex nonlinear relation of traction transformer fault diagnosis. With considering the probability uncertainty and fuzzy uncertainty of the transformer fault character and omen, the proposed fault diagnosis method adopted the belief rule base and evidential reasoning algorithm. Using IEC three-ratio rule as expertise, the training method is no longer dependent on a large number of fault sample data. The RIMER-DGA transformer fault diagnosis model solves the problem of IEC three-ratio fault code missing and improves the accuracy of fault diagnosis. Furthermore the output mode of the confidence coefficient is more effective in describing the mixed fault types.The new method of insulation state assessment of transformer uses entropy weight TOPSIS theory to solve the problem of state maintenance. This method, which overcomes the effect of subjective factor in determining the weight, determines the weight of evaluation index based on entropy theory. The TOPSIS assessment method is introduced for transformer insulation condition assessment. Based on the cuclidcan distance between the sample and standard model, the insulation assessment problem is transformed into a distance problem of vector space. The insulation class of transformer can be evaluated by the methods.Transformer insulation state is a complex nonlinear system. The specific role of oil gas content in the insulation state still can not be fully explained. Predicting insulation state only according to gas content might ignore the interaction between the various gases. The prediction method based on transformer fault state closeness used the theory of gray and ideal point solution. The method predicted fault state closeness data array based on characteristic parameter and three-ratio rules. The transformer state trends can be obtained by the predicted state and traction transformer log. The state trends are meaningful reference for maintenance arrangements.Electrical test vehicle were designed for traction power supply equipment state maintenance. The mobile test system used the transformer insulation fault diagnosis method, state assessment method and state forecasting method. It can efficiently test traction transformer and comprehensively evaluate transformer by online monitoring data, historical information, family data, procedures, guidelines, traction transformers actual working hours, etc in the enterprise MIS. It also can predict maintenance cycle and improve the scientificity of data analysis. In short, it is valuable for the development of traction transformer state maintenance.