基于固有时间尺度分解与多尺度形态滤波的滚动轴承故障特征提取方法

Fault Feature Extraction Method of Rolling Bearing Based on Intrinsic Time-scale Decomposition and Multiscale Morphological Filtering

为了准确提取出滚动轴承的故障特征并对轴承状态进行评估,提出了一种固有时间尺度分解(intrinsic time-scale decomposition,ITD)与多尺度形态滤波相结合的滚动轴承故障特征提取方法。首先,采用ITD方法将滚动轴承故障信号分解成多个固有旋转分量(proper rotation,PR);然后,对比各个PR分量与原始信号的相关性;最后,采用多尺度形态滤波算法对相关性较大PR分量进行滤波降噪,并提取滚动轴承故障特征频率。采用所建立方法对轴承外圈故障和内圈故障实验数据进行分析。结果表明,所提出的故障特征提取方法能够有效抑制噪声,清晰准确地提取出滚动轴承故障特征频率。

The intrinsic time-scale decomposition( ITD)-multiscale morphological filtering( MSMF),a novel fusing approach was proposed to extract fault feature and evaluate bearing condition. Firstly,the failure signal of rolling bearing is decomposed into multiple proper rotation( PR) by using ITD. Secondly,the correlation between the PR components and the original signal was compared. Finally,using MMF to filter and denoise for the large relativity of PR component,and extract the fault feature of the rolling bearing. The experimental data analyzed showed that the method proposed can effectively repress the noise and accurately extract the fault feature frequency of the rolling bearing.