Enhanced Prognosis of Hybrid Systems with Unknown Mode Changes

In this paper, a new model for degradation has been introduced to cover multiple dynamics for prognostics purposes. Firstly, Augmented Global Analytical Redundancy Relations (AGARRs) have been introduced to track system’s health constantly. Whenever an inconsistency appears, the proposed algorithm checks the Mode Change Signature Matrix (MCSM) and decides if inconsistency is due to a change in modes or an existence of a faulty component. Using Mode Dependent Fault Signature Matrix (MD-FSM), a Set of Candidate Faults will be generated and fed into PF part to estimate the actual fault and parameters of the degradation model. Finally, by applying obtained degradation model, Remaining Useful Lifetime (RUL) will be estimated.