利用声信号对滚动轴承进行故障诊断的研究

旋转机械在运行过程中产生的声信号包含了滚动轴承的运行状态信息,且可采用非接触式测量,本文应用它对滚动轴承进行故障诊断。基于morlet小波变换的包络分析对采集的声信号进行降噪及包络处理,然后变换到频域,提取出特征频率并经过转换后作为线性神经网路的输入向量,辨识滚动轴承的状态。实验表明,本方法对滚动轴承故障诊断是有效的。

Diagnosis of rolling element bearing fault using acoustic signal

Acoustic signal generated by a rotating machine is used in the fault diagnosis of roiling element bearing,because it carries information about the working of rolling element bearing and the microphone collecting the signal does not touch the machine.Envelope a- nalysis technique based on the morlet wavelet transform is employed for reducing noise and envelope analysis of collected acoustic signal.Transform the envelope signal to the frequen- cy domain,extract the fault characteristic frequencies and transform it to the fault charac- teristic vectors to be used as the inputs of linear neural networks in order to identify the fault type of rolling element bearing.The experiments’results show that the diagnosis ap- proach in this paper is effective.