The Fault Diagnosis Method Based on Fuzzy Neural Network

In dealing with the problems of fuzzy reasoning and neural network(FNN) methods in pattern recognition, the paper proposes a fault reasoning method based on fuzzy neural network. The proposed FNN identifies the structure and parameters, and initializes and set the rules. The membership function in the process of fuzzification is optimized. Finally, the simulation results prove the proposed method can diagnose the reasons of the fault and the timeliness is greatly improved.