Remaining Useful Life Prediction of Aero-engines by Appropriate Utilization of Multi-sensor Signals

Abstract This paper presents the prediction of remaining useful life (RUL) with appropriate fusion of multi-sensor signals for the aero-engine, which is the heart of an aircraft. With the rapid development of information technology, health condition of one aero-engine is usually monitored with multiple sensors. To properly utilize these multi-sensor condition information for degradation modeling and RUL prediction is one of the key challenges for condition-based maintenance of the whole aircraft. Thus this paper proposes one statistical method based on health indicator (HI) construction and empirical parametric model for aero-engines RUL prediction. The method is validated with run-to-failure data sets of an aircraft gas turbine engine test-bed developed by NASA. Results show that the proposed method can effectively fuse multi-sensor signals to describe the degradation and predict RUL of the aero-engine.