基于共振稀疏分解的滚动轴承早期微弱故障诊断

Fault diagnosis of rolling bearing's early weak fault based on resonance sparse decomposition

传统方法很难对滚动轴承的早期微弱故障进行有效诊断.共振稀疏分解是一种基于多字典库的稀疏分解方法,可以同时分解出滚动轴承故障信号中的瞬态冲击成分及其持续震荡成分(工频及其谐频成分).该方法在对滚动轴承早期微弱故障信号进行自适应滤波降噪(采用Ensemble Empirical Mode Decomposition,EEMD方法)基础上,对处理后的信号进行共振稀疏分解分析,分别构建高、低品质因子小波基函数字典库,并利用形态学分析方法建立信号稀疏表示的目标函数,进而实现对滚动轴承发生故障时具有低品质因子的瞬态故障成分及其他持续振荡高品质因子噪声成分的成功分离.对分离得到的低品质因子信号成分进行包络解调分析,进而得到较好的故障提取特征结果.通过实验验证了所述方法的有效性.

The traditional signal processing method is very hard to diagnose the rolling bearing＇early stage weak fault successfully.The resonance sparse decomposition is a relative new signal processing method based on multiple dictionaries,and it can separate the high Q-factor transient impact component and the low Q-factor sustained oscillation component contained in the rolling bearing fault vibration signal.The vibration signal of rolling bearing＇early weak fault is filtered by the ensemble empirical mode decomposition（EEMD）firstly,then the filtered signal is handled by the resonance sparse decomposition method：Construct the high Q-factor and low Q-factor wavelet base functions dictionaries to match the high Q-factor transient impact component and the low Q-factor sustained oscillation component respectively,then apply envelope demodulation spectrum method to the obtained low Q-factor component and better fault feature result is extracted.The effectiveness of the proposed method is verified through experiment.