A Fault Diagnosis Approach for Roller Bearing Based on SVM and EMD Envelope Spectrum
基于SVM和EMD 包络谱的滚动轴承故障诊断方法

According to the modulation characteristics of roller bearing fault vibration signals, the limitation of the traditional envelope analysis and the situation that it is hard to obtain enough fault samples, a roller bearing fault diagnosis method based on Support Sector Machine(SVM)and Empirical Mode Decomposition (EMD) envelope spectrum is proposed in this paper. Firstly, vibration signals are decomposed into a finite number of intrinsic mode functions (IMFs), then, the ratios of amplitudes in different characteristic frequencies are defined as the characteristic amplitude ratios after the envelope spectra of some IMFs which include the most dominant fault information are obtained; finally, the characteristic amplitude ratios are regarded as the fault features and served as input parameters of SVM classifier to classify working condition and fault patterns of roller bearings. The analysis results demonstrated the effectiveness of the proposed method.