High Voltage Circuit Breaker Fault Diagnosis Based On Neural Fuzzy Petri Nets

The high circuit breakers are for the important task of power system load control and fault protection, with the critical locations in power system. The failure of high voltage circuit breaker tends to endanger the power grid operation,so the study of diagnosis of high voltage circuit breaker fault technology is imperative. It is important to promptly and accurately find out the fault components determine the fault type when the high voltage circuit breaker fails,and to identify the cause of the problem in time to prevent the failure when the high voltage circuit breaker failure symptom appears. In this thesis, with the help of knowledge and reasoning function of Petri nets, put forward neural fuzzy fault diagnosis method of Petri net (NFPN).This method combines the fuzzy theory, Petri net theory and Neural network algorithm in advantage of the field of fault diagnosis, to improve the precision of the weight calculation to library and stability. At the same time improve the forward and reverse reasoning algorithm and the diagnostic efficiency. In this thesis the following works are done in view of the high voltage circuit breaker fault diagnosis.First, combining the fuzzy theory, Petri net theory and Neural network in advantage of the field of fault diagnosis, a fault diagnosis method based on neural fuzzy Petri nets(NFPN) is analysed. Furthermore, the composition, meaning of each element and modeling rules of NFPN are elaborated.Second, learning from the results of previous studies, a forward and backward reasoning algorithm based on NFPN failure prediction and diagnosis is improved.The forward reasoning algorithm through change can make matrix and fault reasoning matrix description. The backward reasoning algorithm represents The reverse fault propagation reasoning matrix and the minimum cut set theory.Third, by analyzing the working principle of the fault type high voltage circuit breaker internal structure, building the model of high voltage circuit breaker fault diagnosis based on NFPN. FMEA method to deal with fault statistics, calculating the confidence threshold change the weights of library rules of credibility.Fourth, for example, in refusing action failure, working principle and fault type resistance strain gauge weighing sensor, for the fault diagnosis of high voltage circuit breaker. Finally,the Labview software was adopted to design a set of high voltage circuit breaker fault forecast and diagnosis system based on NFPN, maintenance and repair of auxiliary maintenance staff better and faster.