Varying Speed Bearing Fault Diagnosis Based on Synchroextracting Transform and Deep Residual Network

Abstract

An intelligent fault diagnosis method is proposed in this study based on Synchroextracting Transform (SET) and deep residual network (DRN) for fault diagnosis of rolling element bearings operating under varying speed condition. Firstly, the bearing condition monitoring (CM) data is processed using SET to obtain the time frequency spectrum graphs as the feature set. The feature set is then used as the input features to train the DRN model. Finally, the trained DRN model is used for an automated bearing fault diagnosis. The classification results show that the proposed method can achieve high recognition accuracy for rolling bearings operating under varying speed conditions.