Prognostic and Health Management of an Aircraft Turbofan Engine Using Machine Learning

An aircraft's turbofan engine is essential to its operation. As with any component of a machine, engine components are subject to degradation and wear over time. This degradation can affect the performance of the engine and may lead to unexpected failures or decreased efficiency. To ensure proper maintenance, predicting its remaining useful life (RUL) is crucial. This paper presents a prediction framework for the RUL of aircraft engines using machine learning (ML) techniques, specifically a Deep Layer Recurrent Neural Network (DL-RNN) model. The proposed method is compared to other ML algorithms, such as Multilayer-Perceptron (MLP), Cascade Forward Neural Network (CFNN), and Nonlinear Auto Regressive Network with Exogenous Inputs (NARX), using two datasets provided by NASA. The results show that the DL-RNN model has better predictive precision compared to the other methodologies.