Fault Diagnosis of Rolling Bearing Based on WT-VMD and Random Forest

Abstract

This paper presents a fault diagnosis method based on wavelet threshold (WT), variational mode decomposition (VMD) and random forest. First, wavelet threshold method is used to preprocess the original vibration signal to reduce noise interference and enhance fault features; then VMD is used to decompose the denoised signal into several intrinsic mode functions (IMFs) and extract the center frequencies of each IMF; finally, these center frequencies will be input into the random forest classifier as feature vectors for fault type identification. Compared with the results of some existing fault diagnosis methods (such as EMD + variable prediction model, VMD + PSO-PNN, VMD + multi-scale permutation entropy + SVM), it is found that this method greatly improves the accuracy of fault identification. The research results show that this method can effectively identify various fault types of rolling bearing.