基于EMD的谱峭度方法在滚动轴承故障检测中的应用

传统谱峭度方法通常采用基于短时傅里叶变换(Short Time Fourier Transform,STFT)的峭度图方法来实现。针对STFT不能保证对瞬态脉冲这种高度非平稳信号最优的分解效果的缺点,提出一种基于经验模式分解(Empirical Mode Decomposition ,EMD)的谱峭度方法。该方法首先利用EMD和Hilbert变换得到信号的时频分布,然后将信号的时频分布按照不同层数分成若干频段,通过计算各频段的峭度值得到相应的峭度图,再根据峭度最大原则选择滤波频段进行带通滤波,最后对滤波信号采用包络分析确定故障信息。实验结果表明：相比传统基于STFT的谱峭度方法,本文方法更能准确的获得轴承加速度信号的故障特征频率信息。

Application of spectral kurtosis approach based on empirical mode decomposition (EMD) in fault detection of rolling element bearings

Traditional spectral kurtosis method is commonly implemented by kurtogram based on short time Fourier transform (STFT). But the STFT does not guarantee the best decomposition effect on the transient pulse such as the highly non-stationary signal. A spectral kurtosis approach based on empirical mode decomposition (EMD) is proposed for above shortcoming. In this method, firstly, EMD and Hilbert transform are used to obtain the signal time-frequency distribution; then the time-frequency distribution was decomposed into several frequency bands according to different layers, and the kurtogram is obtained by calculating the kurtosis of each frequency band; and then the filtering frequency band is selected to bandpass filter according to the maximum kurtosis principle; finally, envelope analysis is used to determine the fault information of the filtered signal. It can be seen from the experimental results that: the more accurate information of fault characteristic frequency with the bearing acceleration signal is obtained compared to the traditional spectral kurtosis approach based on STFT.