局部均值分解在旋转机械复合故障诊断中的应用

Local Mean Decomposition and Its Application to Rotating Machinery Composite Fault Diagnosis

针对旋转机械复合故障振动信号的非平稳特征,开展一种基于局部均值分解(local mean decomposition,LMD)的旋转机械复合故障诊断方法研究。该方法首先通过局部均值分解方法将振动信号分解为若干个PF分量(product function)和一个残余分量之和,然后通过计算各PF分量与原始复合故障信号的相关系数来确定包含故障特征信息的主要成分;最后针对主要成分中的低频分量进行频谱分析从而提取轴的故障特征。针对主要成分中的高频分量采用包络谱分析提取调制故障特征,即提取轴承故障特征。对齿轮箱的轴承、轴复合故障振动信号的分析结果表明了该方法的有效性和可行性。

For the non-stationary characteristic of vibration signal in rotating machinery composite fault diagnosis an analysis method based on LMD（local mean decomposition） was developed. Firstly, the original vibration signal was decomposed into several stationary PF （product fimction） plus a residual component by using of LMD, then the main components in fault signal were determined by calculation of correlation factor of each PF with the original signal. Finally, the features of shaft fault were extracted by Fast Fourier Transform of low-frequency components, and the features of bearing fault were extracted by envelopment spectrum analysis of high frequency componants. The effectiveness of the method is also proved by analysis to composite fault vibration signal of a gear box.