A NEW ROLLER BEARING FAULT DIAGNOSIS METHOD BASED ON AEDE-SVM METHOD AND VMD-SVD

Abstract

This paper presents a new method for roller bearing fault diagnosis based on support vector machine (SVM) with parameters optimized by Adaptive Elitist Differential Evolution method (AEDE).  First, roller bearing acceleration vibration signals are decomposed into function by using Variational Mode Decomposition (VMD) method. Second, initial feature matrices are extracted from there functions by singular value decomposition (SVD) techniques to obtain single values. Thirdly, these values serve as input vector for AEDE-SVM classifier. Experimental results show that the proposed method gives high classification accuracy (100%) and shorter time than other methods.