Research On Extraction And Diagnosis Of Fault Feature Of Vibration Of Rolling Bearing Based On EEMD

Rolling bearing is one of the most widely used parts in varity of rotating machinery, whose running state often directly affects the performance of the whole machine, so the fault diagnosis of rolling bearing has important practical significance and economic value. When the rolling bearing fault, extract some characteristic parameters including the fault information from vibration signals, so as to reveal its fault types, fault location, fault degree, which is a commonly used diagnostic method. Therefore, how to effectively extract the fault features is very important. The fault feature parameters are classified into time domain parameters and frequency domain parameters. The Wavelet Analysis and Empirical Mode Decomposition (EMD) are two kinds of method for extracting common, Ensemble Empirical Mode Decomposition (EEMD) as the optimization algorithm of EMD, has been applied more and more widely in recent years.This paper adopts a set of methods, which based on the EEMD, to extract kurtosis, root mean square value as the representative of the time domain, and to extract gravity frequency, root mean square frequency, frequency standard deviation as the representative of the frequency domain, and using of Support Vector Machine (SVM) for fault identification. Verify that the method has high correct rate in the diagnosis of faults, through compared with the results of EMD decomposition. At the same time, in the time domain and frequency domain, combining multi feature parameter diagnosis correct rate is higher than that of single parameter. Diagnosis of characteristic parameters of time domain and frequency domain feature parameters combination can achieve higher correct rate.