一种无刷直流电动机故障诊断方法研究

A Fault-Detection Method for Brushless DC Motor

提出一种基于小波熵和SOM神经网络的无刷直流电动机故障诊断方法,以无刷直流电动机的霍尔传感器和驱动器故障为研究对象,采用小波分析对故障信号进行分析,并在此基础上通过小波熵对故障特征进行提取,作为故障诊断部分(神经网络)的输入,由训练得到的神经网络对故障进行分类和识别,最后建立无刷直流电动机仿真模型,对该故障诊断方法进行仿真验证。

A fault-detection method for brushless direct current motor was poposed based on wavelet entropy and SOM network.The research object of the fault was Hall sensor and driver of the brushless direct current motor,then the fault was analyzed by using wavelet,and the fault features were extracted by using wavelet entropy. The fault was classified and identified by SOM network. A simulation model was established and the fault-detection method was verified.