基于自适应增强形态滤波的滚动轴承复合故障分离法

Rolling Bearings Based on Adaptive Enhanced Morphological Filtering Composite Fault Separation Method

针对滚动轴承复合故障难以分离的问题,课题组提出了一种自适应多尺度形态滤波分离方法。首先,利用具有提取周期性特征的多尺度形态滤波器和峭度特征能量积(kurtosis feature energy product, KF)提取出一种主要的故障特征分量;然后,利用奇异值分解(singular value decomposition, SVD)降噪方法对提取的故障特征进行降噪处理,增强故障特征;最后,对去噪信号进行迭代筛选分离,得到多个故障特征模式分量。通过仿真信号与异步牵引电机实际故障信号对比实验,结果表明:该方法能够分离复合故障特征,并有效提取噪声干扰下的故障特征信息。该方法滤波效果强于传统方法,具有较好的工程应用价值。

An adaptive multiscale morphological filtering separation method was proposed for the problem that rolling bearing composite faults are difficult to be separated. Firstly, the multi-scale morphological filter with extracted periodic features and the Kurtosis feature energy product(KF) were used to extract a major fault feature component;then, the singular value decomposition(SVD) noise reduction method was used to denoise the extracted fault features and enhance the fault features. Finally, the denoised signal was iteratively filtered and separated to obtain multiple fault pattern components. By comparing the simulated signal and the actual fault signal of the asynchronous traction motor, the results show that the method can separate the composite fault features and effectively extract the fault feature information under the noise interference, and the filtering effect is stronger than that of the traditional method, which has good engineering application value.