Development Of Monitoring And Diagnostic System For Locomotive Bearings

The rolling element bearing is one of the most important units in locomotive and car. Bearing failures would probably bring great suffering to railway transportation and safety. It is of great importance to detect incipient failures in locomotive bearings as early as possible. Therefore, bearing signal analysis using the wavelet transform is examined, and a novel condition monitoring and fault diagnostic system for locomotive bearing is developed. The contributions and conclusions are made as follows:An overview of the mechanism analysis of locomotive bearing faults is presented, which include the cause of the locomotive bearing faults, basic forms, vibration mechanism, vibration model and fault characteristic frequency. Some techniques for fault diagnosis of rolling element bearing are shown, such as time domain analysis, frequency domain analysis, Hilbert transform, wavelet transform and so on.Two methods which applied wavelet analysis to fault diagnosis is proposed, which are the noise reduction analysis method using wavelet transform and feature extraction analysis method by wavelet packet transform. Based on the theory of wavelet analysis, the wavelet transform is used to reduce the noise of the vibration signal of bearings. Furthermore, the fault character frequency of roller bearings and the energy character of lowest crunodes are extracted by means of wavelet package multiplayer decompose and reconstruction. The effectiveness and practicability of the proposed methods is verified by the computer generated signal and the vibration experiment on bearing test rig.Based on the investigations mentioned above, a monitoring and diagnostic system for locomotive bearings is developed The system is successful applied to the bearings of the running gear of a locomotive. The effectiveness of proposed system is verified by stationary test.