Fault Diagnosis of Rolling Bearing Based on ICEEMDAN and SSA-RVM

Abstract

Abstract In order to effectively improve the fault identification rate of correlation vector machine (RVM) in bearing fault model diagnosis and identification, a rolling bearing troubleshooting method improvement-based adaptive white noise complete integrated empirical mode decomposition (ICEEMDAN) and sparrow search algorithm (SSA) optimized correlation vector machine (RVM) is proposed. A rolling bearing troubleshooting method improvement-based adaptive white noise complete integrated empirical mode decomposition (ICEEMDAN) and sparrow search algorithm (SSA) optimized correlation vector machine (RVM) was presented. First, the ICEEMDAN method is used to decompose the raw vibration signal, and the power entropy and the proportion of the power of the intrinsic mode component (IMF) to the total power of the signal are taken as the fault characteristics. Optimization of the width factor and hyperparameters of RVM using SSA algorithm, and the RVM model is built to realize fault Identification of rolling bearings. The test results indicate that the SSA-RVM model has a high fault identification rate and dramatically improves the accuracy of fault diagnosis.