Fault diagnosis of rolling bearing based on EMDD information and KNP-SVDD

CHEN Yu-chen, HE Yi-bin, DAI Qiao-sen, LIU Xiang, HE Su-xun
(Mechanical and Electrical Engineering, Wuhan Institute of Technology, Wuhan 430205, China)

Abstract: Aiming at the problem of poor diagnosis effect when all kinds of data samples were unevenly distributed in fault diagnosis, support vector data description (SVDD) was proposed based on support vector machine (SVM), the extension of SVDD to multiple decision and the limitations of various extension methods were also studied. A multi-decision SVDD weighted by K-neighbor probability was proposed. The ensemble empirical mode decomposition (EEMD) was used to decompose the original signal, and the information content of each intrinsic mode function (IMF) was calculated and taken as a characteristic. The third-party experimental data were used to test the identification accuracy of k-neighbor probability support vector data description (KNP-SVDD) method in various fault categories. The results indicate that the method can effectively identify the location and severity of the fault, and the superiority of the method is proved by comparing with other classification methods.

Key words: fault diagnosis; support vector machine（SVM）; ensemble empirical mode decomposition（EEMD）; intrinsic mode function（IMF）; K-neighbor probability support vector data description（KNP-SVDD）