基于FIR-EMD和改进SVM的铁路轴承故障诊断

Fault Diagnosis of Railway Bearings Based on FIR-EMD and Improved SVM Algorithm

针对铁路轴承故障难以有效识别的问题,提出了一种基于FIR-EMD和改进SVM的铁路轴承故障诊断方法。首先,对采集到的振动信号进行FIR降噪,再对降噪后的信号进行EMD分解,接着对分解后的信号构造IMF能量矩,最后将能量矩作为改进SVM的输入实现铁路轴承故障分类。实验结果表明该方法能有效地识别铁路轴承故障类型。

It is difficult to effectively identify the faults in railway bearings.In this paper,a scheme for railwaybearings fault diagnosis based on FIR-EMD and improved SVM algorithm is proposed.First of all,the signal is processedthrough the signal de-noising based on the FIR.Then,the de-noised vibration signals of the railway bearings aredecomposed by EMD,and the IMF energy moments are calculated through the decomposed vibration signals.Finally,theIMF energy moments are used as feature vectors and input into the improved SVM to realize faults classification for therailway bearings.The results of an example show that the proposed diagnosis approach can effectively identify the faultpatterns of railway bearings.