基于EEMD和快速谱峭度的滚动轴承故障诊断研究

Fault diagnosis of rolling bearing based on EEMD and fast spectral kurtosis

从经验模态分解(EMD)中得到的本征模态函数(IMF)存在模态混合现象,并且其末端效应影响分解效果,针对这一问题,以西安交通大学XJTU-SY滚动轴承中的振动数据为例,对其外圈振动的正常数据和失效数据分别进行了对比研究。首先,采用集合经验模态(EEMD)分别对正常和失效轴承的振动信号进行了分解,得到了各阶IMF分量;然后,通过峭度准则选择关键的IMF分量进行了信号重构,计算了重构的信号快速谱峭度,根据快速谱峭度得出的中心频率和带宽为依据,对重构信号进行了带通滤波处理;最后,对包络谱进行了对比分析,获得了滚动轴承的准确故障特征信息。研究结果表明:通过EEMD分解和快速谱峭度得出滤波后的重构信号降噪效果明显,可以得到良好的故障带宽和中心频率;该方法能有效测出XJTU-SY滚动轴承出现外圈故障时的振动频率。

Aiming at the problem of modal mixing and end effects affecting the decomposition effect in the intrinsic mode function(IMF)obtained by empirical mode decomposition(EMD),the vibration data of XJTU-SY rolling bearing in Xi’an Jiaotong University was token as an example,and the normal data and failure data of the outer ring vibration were compared an analyzed respectively.First,the normal and failed vibration signals were decomposed by the ensemble empirical mode decomposition(EEMD),and each order of IMF components was obtained.Then the key IMF components were selected to reconstruct the signal through the kurtosis criterion,and the fast spectral kurtosis of the reconstructed signal was calculated.Based on the center frequency and bandwidth obtained by fast spectral kurtosis,the reconstructed signal was processed by bandpass filtering.Finally,the envelope spectrum was compared and analyzed,and the accurate fault characteristic information of the rolling bearing was obtained.The results show that the noise reduction effect of the filtered reconstructed signals obtained by EEMD decomposition and fast spectral kurtosis is obvious,and good fault bandwidth and center frequency can be obtained.This method can effectively measure the vibration frequency of XJTU-SY rolling bearing when the outer ring fault occurs.