Combined EMD-SVM technique for fault diagnosis of roller bearing
Accurate diagnosis of faults roller bearing can significantly enhance the reliability and safety of mechanic systems.This paper describes a new approach using empirical mode decomposition (EMD) for extraction of features from roller bearings and support vector machine to classify the patterns inherent in the features extracted through the EMD of different fault patterns.The experimental results from roller bearing signals with faults show that the diagnosis approach based on support vector machine (SVM) by using EMD to extract the energy of different frequency bands as features can identify roller bearing fault patterns accurately and effectively and is superior to that based on LVQ neural newwork (LVQNN) and RBF neural newwork (RBFNN) at each training set size.The Gaussian RBF kernel has been used for training and testing of the SVM and the values of the variance of the Gaussian and punitive parameter (C) have been chosen as 0.04 and 2.54× 104,respectively.The analysis results also show that he proposed combined EMD-SVM technique has a robust anti-noise performance in a noisy environment.The application in fault diagnosis of roller bearing shows that the proposed EMD-SVM technique has strong practicability.
Empirical mode decomposition Support vector machine Fault diagnosis
Xian Guangming
Computer Engineering Department of Nanhai Campus, South China Normal University, Guangdong Foshan,China, 528225
国内会议
山东泰安
英文
546-549
2009-08-15（万方平台首次上网日期，不代表论文的发表时间）