A Fusion Prognostic Method for Remaining Useful Life Prediction Based on an Extended Belief Rule Base and Particle Filters

As a critical part of prognostics and health management (PHM), remaining useful life (RUL) prediction can provide manufacturers users with system lifetime information improve the reliability maintainable systems. Particle filters (PFs) are powerful tools for RUL because they represent uncertainty results well. However, due to lack measurement data, parameters model cannot be updated during long-term process. Additionally, complex systems, often obtained in an analytical form. In this paper, fusion prognostic method based on extended belief rule base (EBRB) PF is designed solve these problems. proposed framework, double-layer maximum mean discrepancy-extended (DMMD-EBRB) time delay adopted estimate predict hidden behavior degrading system. The unknown degradation identified by using output EBRB. Afterwards, state further predicted PF. effectiveness validated NASA-PCoE CALCE lithium-ion battery experiment datasets. addition, several other related methods investigated comparison method. experiments show that yields better performance than existing methods.