Mechanical fault diagnosis method of 10kV distribution network high voltage circuit breaker based on neural network
Because the fault characteristics of high-voltage circuit breakers are multi-source heterogeneous data,a single fault characteristic can not accurately characterize the fault characteristics of equipment,resulting in low fault diagnosis accuracy.Therefore,a mechanical fault diagnosis method of 10kV distribution network high voltage circuit breaker based on neural network is proposed.The vibration signal of the equipment is collected and processed by subsection and discrete cosine transform,and the energy entropy of the signal frequency band is obtained.The mechanical fault feature quantity is selected by combining amplitude-frequency gradient,and the priority of the fault feature quantity is determined by combining Fourier transform method,and the fault feature space is constructed,and then the single fault feature is fused.Based on this,the matching degree of the test sample relative to a certain fault category is calculated,so as to determine the fault type of the sample.The experimental results show that the mechanical fault types of equipment obtained by the proposed method are completely consistent with the actual fault types,and the fault diagnosis accuracy is high.
fault characteristicsfault signal amplitude and frequencyneuronmechanical failurefault matching degree
万方数据
维普
