The research on fault diagnosis of rolling bearing based on current signal CNN-SVM
Xinghua Wang, Runxin Meng, Guangtao Wang, Xiaolong Liu, Xiaohong Liu, Daixing Lu
Abstract
Abstract This article proposes a novel approach to address the issues of low accuracy in fault diagnosis and the difficulty in installing sensors on rolling bearings in mechanical and electrical equipment systems. To accomplish fault diagnosis of rolling bearings, a network structure algorithm based on convolutional neural network (CNN) and support vector machine (SVM) is presented, which incorporates the electric motor current signal. Firstly, the collected electric motor current signal is subjected to a wavelet filter with a soft-hard threshold to eliminate the noise. Secondly, the processed data is fed as input to a one-dimensional CNN to perform feature extraction and dimensionality reduction. Finally, the dimensionality-reduced features are processed by a SVM to diagnose rolling bearing faults. The research results indicate that the proposed method significantly improves the accuracy of rolling bearing fault diagnosis compared to other approaches, with an accuracy of up to 99.01%. This study introduces an innovative approach that can be applied to the field of rolling bearing fault diagnosis, offering valuable insights for research and application in this domain.