Identifying new prognostic features for remaining useful life prediction using particle filtering and Neuro-Fuzzy System predictor

An accurate prediction of the remaining useful life (RUL) from a prognosis system relies on a good selection of prognosis features. The latter should well capture the trend of the fault progression. In situation where the development of degradation model is difficult, we must be addressed to the identification of new features having an obvious trending quality. in this context, This paper present a new selection method based upon a Particle Swarm Optimization algorithm to identify the advanced prognosis feature and a particle filtering for the prediction of the remaining useful life. The fault growth model is integrated to the particle filter using a Neuro-Fuzzy System with its process noise. This method was validated on a set of experimental data collected from bearings run-to-failure tests.