Time Series Multi-Channel Convolutional Neural Network for Bearing Remaining Useful Life Estimation

This paper proposes a deep learning method, called time series multi-channel convolutional neural network (TSMC-CNN), for remaining useful life (RUL) estimation of bearings. The time series data of bearing operation are divided into multiple channels to be fed into the convolutional neural network (CNN) to extract relationship between far apart data points. The PRONOSTIA bearing operation datasets are used to evaluate the proposed method performance. The evaluation results are compared with those of related methods to show the superiority of the proposed method in terms of the root mean squared error (RMSE) and the mean absolute error (MAE).