基于WAEEMD和MSB的滚动轴承故障特征提取

针对调制信号双谱（MSB）方法仅能处理平稳信号的不足，提出了一种基于加权平均集成经验模态分解（WAEEMD）和MSB的滚动轴承故障特征提取方法。首先，利用WAEEMD将滚动轴承的非平稳振动信号分解成一系列具有平稳特性的固有模态函数（IMF）;然后，开发了一种基于Teager能量峭度（TEK）的加权平均方法以强调敏感IMF的重要性，并将加权后的IMF重构为WAEEMD滤波信号;最后，应用MSB分解WAEEMD滤波信号中的调制分量并提取故障特征频率。仿真和实验结果表明，相对于快速谱峭度（FK）和EEMD-MSB方法，WAEEMD-MSB方法能更准确地获取故障特征，从而验证了WAEEMD-MSB方法的有效性。

Feature Extraction of Rolling Bearings Based on WAEEMD and MSB

Aiming at the facts that the modulation signal bispectrum(MSB) might only process stationary signals, a novel method for fault feature extraction of rolling bearings was proposed based on the WAEEMD and MSB. Firstly, vibration signals of rolling bearings were decomposed into a list of intrinsic mode functions(IMFs) by ensemble empirical mode decomposition(EEMD). Subsequently, the IMFs were reconstructed into the WAEEMD filtered signals using the weighted average method based on Teager energy kurtosis (TEK). Finally, the MSB was used to decompose the modulated components in the WAEEMD filtered signals and extract the fault characteristic frequencies. The analysis results illustrate that the WAEEMD-MSB has a superior performance over fast kurtogram (FK) and EEMD-MSB in extracting bearing fault features.