基于局部特征尺度分解与最小熵解卷积的轴承故障诊断

Fault Diagnosis for Bearings Based on LCD-MED

为准确进行滚动轴承的故障诊断,结合局部特征尺度分解(LCD)和最小熵解卷积(MED)给出了一种新的故障诊断方法。首先,对轴承振动信号进行局部特征尺度分解,得到若干个内禀尺度分量;然后,依据互相关系数指标,采用聚类分析方法自动选取有用分量并叠加作为重构信号;最后,应用最小熵解卷积将重构信号降噪,并应用包络分析技术进行故障诊断。通过轴承内、外圈故障振动数据的分析表明:经LCD-MED处理后,振动信号的峭度值得到了较大提高,故障特征频率更加突出,基于LCD-MED的方法在轴承故障诊断中有效且合理。

A new fault diagnosis method is proposed based on local characteristic-scale decomposition（ LCD） and minimum entropy deconvolution（ MED） to accurately diagnose of fault of rolling bearings. Firstly,the vibration signal of bearings is decomposed into some intrinsic mode components（ ISCs） by LCD. Secondly,according to cross correlation coefficient index,the useful component is selected automatically and superposed as reconstructed signal by using cluster analysis method. Then the reconstructed signal is denoised by MED,and the fault is diagnosed by envelope analysis technology. The analysis of fault vibration data of inner and outer rings of the bearings shows that the kurtosis of vibration signal is improved greatly and the frequency of fault features is highlighted after LCD-MED treatment. The method based on LCD-MED is effective and reasonable in fault diagnosis of the bearings.