Rolling Element Bearings Fault Intelligent Diagnosis Based on Convolutional Neural Networks Using Raw Sensing Signal

Abstract

Vibration signals captured by the accelerometer carry rich information for rolling element bearing fault diagnosis. Existing methods mostly rely on hand-crafted time-consuming preprocessing of data to acquire suitable features. In contrast, the proposed method automatically mines features from the RAW temporal signals without any preprocessing. Convolutional Neural Network (CNN) is used in our method to train the raw vibration data. As powerful feature exactor and classifier, CNN can learn to acquire features most suitable for the classi_cation task by being trained. According to the results of the experiments, when fed in enough training samples, CNN outperforms the exist methods. The proposed method can also be applied to solve intelligent diagnosis problems of other machine systems.