关键词: 广义形态滤波, 多分辨奇异值分解（MRSVD）, Hilbert包络谱分析, 故障诊断

Abstract: In order to obtain the fault information of rolling bearing under complicated working conditions, a method based on generalized morphological filter and Multi-Resolution Singular Value Decomposition（MRSVD） for fault diagnosis is proposed in this paper. Firstly, the bearing vibration signals are de-noised by using generalized morphological filter. Then the filtered signals are decomposed into a series of component signals by MRSVD. Finally, the detail signals which contained the most valuable fault features are selected by kurtosis criterion. The Hilbert envelope spectrum analysis is applied to the details signals afterwards to obtain the fault information. The experimental results have shown that the proposed method is capable of extracting useful fault features from complex vibration signals, and has offered an approving performance on fault diagnosis of the rolling bearing.

Key words: generalized morphological filter, Multi-Resolution Singular Value Decomposition（MRSVD）, Hilbert envelope spectrum analysis, fault diagnosis