Fault diagnosis and remaining useful life estimation of aero engine using LSTM neural network

Hero engine is a kind of sophisticated and expensive industrial product. Accurate fault location and Remaining Useful Life (RUL) estimation for aero engine can lead to appropriate maintenance actions to avoid catastrophic failures and minimize economic losses. The aim of this paper is to propose utilizing Long Short-Term Memory (LST NI) neural network to get good diagnosis and prediction performance in the cases of complicated operations, hybrid faults and strong noises. The whole proposition is demonstrated and discussed by carrying out tests on a health monitoring dataset of aircraft turbofan engines provided by NASA. Performances of LSTM and some of its modifications were tested and contrasted. Experiment results show the standard LSTM outperformed others.