基于局部均值分解和最大相关峭度解卷积的滚动轴承早期故障提取

Early Fault Extraction of Rolling Bearing based on LMD and MCKD

滚动轴承处于早期故障阶段的时候,特征信号比较微弱,同时受到干扰噪声的影响,造成故障特征难以提取。针对这一问题,提出了基于局部均值分解(Local mean decomposition,LMD)和最大相关峭度解卷积(Maximum correlated kurtosis deconvolution,MCKD)两者相结合的故障诊断方法。在强噪声背景条件下,LMD对微弱故障信号特征难以提取,故对LMD分解得到的一组乘积函数(Product function,PF),利用相关系数与峭度值筛选出敏感分量进行信号重构,然后利用MCKD进行滤波,突出故障信号尖脉冲,最后根据信号的包络功率谱提取故障特征频率。算法仿真和实验证明了该方法的有效性。

When the roller bearings are in the early stage of failure,the characteristic signal is weak and it is affected by the interference noise,which makes the fault feature difficult to extract. In order to solve this problem,a fault diagnosis method based on the combination of local mean decomposition( LMD) and maximum correlated kurtosis deconvolution( MCKD) is proposed. Under the strong noise environment,LMD is difficult to extract the characteristic of weak fault signal,therefore,a set of PF components that are decomposed by LMD which use the correlation coefficient and kurtosis value select sensitive component for signal reconstruction. Then MCKD is used to filter and it is used to highlight the pulse of fault signal. Finally,according to the signal envelope power spectrum,the fault characteristic frequency is extracted,the effectiveness of the method is proved by some simulation and application examples.