Prediction of Bearing Remaining Service Life Based on CNN-LSTM

Abstract

Aiming at the waste of resources caused by the bearing reaching the service time and still meeting the service conditions, a bearing remaining service life prediction method based on CNN-LSTM is proposed. Firstly, a high-speed railway traction motor bearing which has completed service but is still healthy is selected as the research object, the test platform is built and the bearing vibration signal is collected; secondly, a network model of CNN-LSTM is established; then, the collected vibration signal is input into the network model after Fourier transform, and its deep features are mined; finally, the remaining service life is predicted through the prediction module. The results show that the predicted value obtained by the proposed method is closer to the true value, which can well reflect the performance degradation trend of the bearing in operation.

Keywords

• Rolling bearing;CNN-LSTM;Remaining service life prediction;Long and short term memory network