ABSTRACT
To monitor the railway axle bearing running statues online efficiently and accurately, a novel method based on target entropy scale determination was proposed for the fault diagnosis of the axle bearing. The target entropy scale was determined by four information entropies based on EEMD. Through wheel running test rig, the vibration signals of eight types of the statuses of railway axle bearings were collected containing the normal condition, single faults condition and multiple coupling faults condition. In the light of the proposed fusion information entropy method, the diagnosis of railway axle bearing faults was completed. The results show that the fusion entropy method based on EEMD holds high precision in the recognition of axle bearing faults. The efforts of this study provide a promising and useful approach for the process fault diagnosis of the railway axle bearing.
