ELMD和MCKD在滚动轴承早期故障诊断中的应用

Early Fault Diagnosis of Roller Bearing based on Ensemble Local Mean Decomposition and Maximum Correlated Kurtosis Deconvolution

针对滚动轴承早期故障特征信号微弱且受环境噪声影响严重,故障特征信息难以识别的问题,提出了基于总体局部均值分解(Ensemble local mean decomposition,ELMD)和最大相关峭度反褶积(Maximum correlated kurtosis deconvolution,MCKD)的早期故障诊断方法。该方法首先运用ELMD对采集到的振动信号进行分解,得到有限个乘积函数(Product function,PF),由于噪声的干扰,从PF分量的频谱中很难对故障做出正确的判断。然后对包含故障特征的PF分量进行最大相关峭度反褶积处理以消除噪声影响,凸现故障特征信息。最后对降噪信号进行Hilbert包络谱分析,即可从中准确地识别出轴承的故障特征频率。通过轴承故障模拟实验和工程应用实例验证了该方法的有效性与优越性。

Aiming at the problem that early fault characteristic signal of rolling bearing is weak and affected by the environmental noise seriously,which makes the fault feature information difficult to be identified,a rolling bearing early fault diagnosis method is proposed in this study,based on ensemble local mean decomposition and maximum correlated kurtosis deconvolution. First of all,the ensemble local mean decomposition was adopted to decompose collected vibration signals,and obtains a finite number of product functions. Due to the interference of noise,it is difficult to make a correct judgment of the fault from the spectrum of the PF component. Then,the PF component that contains fault feature was processed by using maximum correlated kurtosis deconvolution to reduce the strong background noise and enhance the fault information. Finally,the fault frequency can be obtained accurately by Hilbert envelope spectrum. The bearing experimental signal and actual engineering data analysis has verified the effectiveness of the proposed method.