基于谱峭度和CEEMD的滚动轴承声信号故障诊断研究

使用声信号来诊断轴承故障越来越受到重视。针对滚动轴承故障信号的强背景噪声特点,提出一种基于谱峭度和互补集合经验模态分解(CEEMD)的故障特征提取方法。该方法首先对滚动轴承声信号进行快速谱峭度计算并进行带通滤波预处理,使滚动轴承声信号变得简单且噪声小,故障冲击成分明显;然后利用CEEMD将滤波信号进行分解运算,得到一系列本征模态(IMF)分量;再利用相关系数法和时域特征指标峰值因子选取包含故障信息最丰富的IMF分量;最后用Hilbert算法包络解调分析选取的IMF分量,得到清晰的故障特征频率。经滚动轴承故障实验分析,该方法可以对滚动轴承故障进行有效的诊断。

Research on fault diagnosis of acoustic signals of rolling bearings based on spectral kurtosis and CEEMD

The use of acoustic signals to diagnose bearing fault is receiving increasing attention.Aiming at the strong background noise characteristics of rolling bearing fault signal,a fault feature extraction method based on spectral kurtosis and Complementary Ensemble Empirical Mode Decomposition(CEEMD)was proposed.This method firstly performs fast spectral kurtosis calculation on the acoustic signal of the rolling bearing and preprocesses the band-pass filter,so that the acoustic signal of the rolling bearing becomes simple and the noise is small,and the fault impact component is obvious;the CEEMD is used to decompose the filtered signal to obtain a series of Intrinsic Modal Function(IMF)components;then the correlation coefficient method and the time domain feature index peak factor are used to select the IMF component that contains the most abundant fault information;finally,the selected IMF components are analyzed by Hilbert algorithm envelope demodulation to obtain clear fault characteristic frequencies.Based on the analysis of rolling bearing failure test,this method can effectively diagnose rolling bearings fault.