基于IMCKD与谱负熵的滚动轴承故障诊断方法

Fault diagnosis method for rolling bearings based on IMCKD and spectral negentropy

针对轴承故障成分常以周期性冲击成分出现在振动信号中,而冲击信号常被强大噪声淹没,导致轴承故障诊断难度较大的问题,提出一种基于改进的最大相关峭度解卷积(IMCKD)与谱负熵的故障特征提取方法。首先,采用改进的最大相关峭度解卷积以最大相关峭度为目标对原始振动信号进行降噪处理,检测信号中的周期性冲击成分;然后,以最大谱负熵值为准则寻找信号的最佳分析频段;最后,通过平方包络解调提取出轴承的故障特征。仿真和实测信号验证了该方法的有效性。

As components of bearing fault often appear in the vibration signal as periodic impulse components which are often submerged by strong noise, the extraction of periodic impulse can be very difficult. A new method is proposed to solve the problem based on improved maximum correlation kurtosis deconvolution (IMCKD) and spectral negentropy. Firstly, the IMCKD is used to denoise the original vibration signal with the maximum correlation kurtosis as the target. And then, the spectral negentropy is used as the criterion to find the best analysis frequency band. Finally, the fault feature of the bearing is extracted through the square envelope analysis. The validity of the method is verified by simulation and measured signals.