基于Gabor变换降噪和盲信号分离的轴承故障诊断方法

Fault diagnosis of rolling bearings based on Gabor transform denoising and blind signals separation

针对轴承故障信号往往淹没在强烈的系统噪声中导致故障特征难以提取的情况,提出了一种基于Gabor变换降噪和盲信号分离的故障诊断方法。该方法利用Gabor变换对时频信号的优良分辨率和盲信号分离技术的优势,先对非平稳信号进行降噪,再通过盲信号分离技术对降噪后的信号进行分离,提取出故障频率。实验结果表明,该方法能有效地诊断出轴承故障特征。

Early fault signals of a rolling bearing submerged in system noises is weak and difficult to identify .Fault diagnosis of rolling bearings based on Gabor transform denoising and Blind Signals Separation （ BSS） is proposed .Using the Gabor transform with good time-frequency resolution and the advantages of BSS for rolling bearing fault diagnosis ,the composite noise of system signals was eliminated with the Gabor transform algorithm ,the denoised signals were separated with the BSS algorithm ,then the fault signal characters were extracted .The experimental results showed that this method is successful in separating fault patterns for rolling bearing .