The Application of the Fuzzy Neural Network-Wavelet Singularity Detection in Mechanical Fault Diagnosis of High Voltage Breakers

In this paper, a new method that the wavelet singularity detecting and NN of fuzzy combined to process the mechanism libration signal of the HV breaker is introduced, which is called the NN of fuzzy singularity. This is a new method that the wavelet singularity exponent of fuzzification combined with the multilayer feedforward NN is imported to the mechanism fault diagnosis of the HV breaker. The experimental results show that this method can achieve better effect than the wavelet singularity detecting and improves the accuracy and precision of diagnosis