加权多尺度形态滤波在轴承故障诊断中的应用

摘要

Vibration signals of defective rolling bearings is often characterized by the presence of pe?riodic impulsive attenuation components,and these impulse features could be extracted utilizing multi-scale morphological filter.To distinguish the contribution of analyzed results on different scales, an index is put forward to evaluate the feature extraction performance,which is then adopted for the weighted summation of multi-scale analyzed results.The proposed method is evaluated by both simulated impulsive signal and vibration signal measured on defective bearings with outer race fault, respectively.Results show that the weighted multi -scale morphological filter method has the superior performance in extracting impulsive component, especially for the high level noise signals,which demonstrates that the method is an efficient tool for bearing fault diagnosis.%滚动轴承故障信号中通常包含了周期性出现的冲击衰减成分,冲击频率反映了轴承发生故障的位置信息.多尺度形态滤波可以提取分布在不同尺度下的冲击特征,为了更准确的反映不同尺度分析结果的贡献,提出了一种评价冲击特征提取效果的指标,并以此计算权值对多尺度分析结果进行加权综合.仿真信号和轴承外圈故障信号的分析结果表明,加权多尺度形态滤波方法可以有效提取强背景噪声下的冲击特征,为滚动轴承故障诊断提供了一种有效的手段.