Fault Diagnosis of Gear Using Oil Monitoring Samples and Vibration Data

The objective of this work is to put forward a method of machine fault diagnosed fusing oil monitoring data and vibration data. Firstly, after analyzing the meanings of oil monitoring samples and vibration data, the fusion method in decision-making layer, which is Dempster-Shafer theory of evidence, is chosen in the paper. Then, a new kind of fault diagnosis method fusing oil monitoring samples and vibration data based on Dempster-Shafer theory of evidence is established. This method includes two main steps. The first step is to get individual diagnosis result by using oil monitoring samples data or vibration data separately. The method is BP neural network. The second step is to fuse two different results to get a final result. The method is Dempster-Shafer theory of evidence. In the paper, fault diagnosis of the gear is used to verify the method above. These oil monitoring data and vibration data are collected from several experiments in which gear wear is simulated. In the end, the fusion result is compared with two single results, and the analysis shows that the proposed method is more effective than single method.