A Hybrid Data-Driven Approach for Predicting Remaining Useful Life of Industrial Equipment

Abstract

Guaranteeing the safety of equipment is extremely important in industry. To improve reliability and availability of equipment, various methods for prognostics and health management (PHM) have been proposed. Predicting remaining useful life (RUL) of industrial equipment is a key aspect of PHM and it is always one of the most challenging issues. With the rapid development of industrial equipment and sensing technology, an increasing amount of data on the health level of equipment can be obtained for RUL prediction. This paper proposes a hybrid data-driven approach based on stacked denoising autoencode (SDAE) and similarity theory for estimating remaining useful life of industrial equipment, which is named RULESS. Our work is making the most of stacked SDAE and similarity theory to improve the accuracy of RUL prediction. The effectiveness of the proposed approach was evaluated by using aircraft engine health data simulated by commercial modular Aero-Propulsion system simulation (C-MAPSS).