Abstract

Failure prognosis and uncertainty representation in longterm predictions are topics of paramount importance when trying to ensure safety of the operation of any system. In this sense, the use of particle filter (PF) algorithms-in combination with outer feedback correction loops- has contributed significantly to the development of a robust framework for online estimation of the remaining useful equipment life. This paper explores the advantages of using a combination of PF-based anomaly detection and prognosis approaches to isolate rare events that may affect the understanding about how the fault condition evolves in time. The performance of this framework is thoroughly compared using a set of ad hoc metrics. Actual data illustrating aging of an energy storage device (specifically battery state-of-health (SOH) measurements [A-hr]) are used to test the proposed framework.