The signal characteristics of substation switch faults have a wide range,and there are problems of con-fusion and unclear characteristics.This paper presents a method of monitoring substation switch fault conditions based on sound signals.The probability density function is used to calculate the abnormal operation probability of substation switchgear and determine the key monitoring switchgear.Pre-emphasis and frame processing are carried out for sub-station switchgear signals,and a Mel filter is established to extract MFCC features of substation switchgear signals.MFCC features are input into the learning vector neural network(LVQ)to output the fault state monitoring results,and a backtracking search algorithm is introduced to optimize the initial weight of the learning vector neural network to a-chieve the substation switch fault state monitoring.The simulation results show that the proposed method has good fault condition monitoring effect.
关键词
变电站开关/概率密度函数/学习矢量神经网络/梅尔滤波器/故障状态监测
Key words
Substation switch/Probability density function/Learning vector neural network/Meier filter/Fault condition monitoring
维普
万方数据