Application of Weighted Degree of Grey Incidence of Optimized Entropy and KFCM for Fault Diagnosis of Circuit Breaker

With the development of power grid, electrical equipment is required to be more and more intelligent. In this paper, a fault diagnostic method combining weighted degree of grey incidence of optimized entropy algorithm and Kernel Fuzzy Cluster Method(KFCM) is proposed. By extracting characteristic values of the current signals of different main fault types, fault database of the current signals can be established. KFCM is utilized to train fault samples of current signals to form the reference sequence. The testing data is regarded as the comparison sequence. Meanwhile, weighted degree of grey incidence of optimized entropy algorithm is utilized to calculate the correlation degree between the two sequences. Finally the fault type of circuit breakers is identified based on the correlation degree. Experiments have proved that this method achieves perfect results in diagnosing main mechanical faults of circuit breakers.