Fault diagnosis of rolling bearing based on LCD and kurtosisenergy ratio criterion

YANG Wen-zhi, ZHANG Ru-jun, AN Wen-bin

(College of Mechanical Engineering, Inner Mongolia University of
Science and Technology, Baotou 014010, China)

Abstract: Aiming at the problem of unscreened effective components in the fault diagnosis of rolling bearings in the local feature scale decomposition, the intrinsic scale component(ISC) generated by LCD decomposition was screened through the horn-to-energy ratio criterion. A method based on LCD decomposition and cradle-energy ratio criteria was proposed. First, the collected rolling bearing vibration signal was decomposed by LCD to obtain the ISC components of different energies, and the valid ISC components were screened by using the horn-energy ratio criterion, and the energy entropy and multi-scale entropy of the valid ISC components after screening were calculated. The result of the calculation was combined to construct the feature vector. Finally, the fault diagnosis of the rolling bearing was realized through the support vector machine(SVM) fault classifier. The results indicate that the redundant component of effective component is reduced by using the skewenergy ratio criterion, and the fault diagnosis accuracy of the rolling bearing is obviously improved in inner and outer ring fault diagnosis.

Key words: local feature scale decomposition（LCD）; rolling bearing; fault diagnosis; energy entropy; multiscale entropy; feature fusion