Probabilistic forecasting informed failure prognostics framework for improved RUL prediction under uncertainty: A transformer case study

Jose Ignacio Aizpurua, Brian G. Stewart, Stephen D. J. McArthur, Markel Penalba, Manex Barrenetxea, Eñaut Muxika, John V. Ringwood

Abstract: Highlights•Probabilistic forecasting and experimental lifetime models informed failure prognostics framework.•Probabilistic forecasting integrated with Bayesian state-estimation models.•Probabilistic remaining useful life (RUL) estimation under uncertainty.•RUL estimates quantify confidence on the experienced operational conditions.•Two industrial case studies for transformers operated in different power plants.