Remaining useful life estimation using long short-term memory deep learning

Abstract

This paper proposes a deep learning method to estimate the remaining useful life (RUL) of aero-propulsion engines. The proposed method is based on the long short-term memory (LSTM) structure of the recurrent neural network (RNN). LSTM can effectively extract the relationship between data items that are far separated in the time series. The proposed method is applied to the NASA C-MAPSS data set for RUL estimation accuracy evaluation and is compared with the methods using the multi-layer perceptron (MLP), support vector regression (SVR), relevance vector regression (RVR) and convolutional neural network (CNN). Comparisons show that the proposed method is better than others in terms of the root mean squared error (RMSE) and the value of a scoring function.