SGBRT: An Edge-Intelligence Based Remaining Useful Life Prediction Model for Aero-Engine Monitoring System

In this paper, we develop an edge intelligence based aero-engine performance monitoring system. The proposed approach can effectively predict the remaining useful life of aero-engines, which is the main focus within the prognostics and health management framework - thus it provides support for optimal operation planning and maintenance decisions. The proposed model, which we term SGBRT, follows a hybrid machine learning approach, combining a self-organizing mapping network with a gradient boosting regression tree model. In particular, the SGBRT computes the remaining useful life of an aero-engine in two steps: it first employs a self-organizing map to cluster the sample data; and then it fits each cluster by way of a gradient boosting regression tree. Detailed simulation results with the C-MAPSS dataset show that this method achieves a higher prediction accuracy and better generalization than other conventional approaches; the compared methods range from classical approaches such as a switching Kalman filter to state-of-the-art deep learning models.