Remaining useful life estimation using accelerated degradation test, a gamma process, and the arrhenius model for nuclear power plants

Abstract

Predictive maintenance can be performed in nuclear power plants to prevent the failure of their components and equipment. Estimating the remaining useful life (RUL) of the components and equipment is the most important step in a predictive maintenance strategy. In this study, we focus on the RUL estimation of bipolar junction transistors (BJTs). We propose a practical method that applies a gamma process and the Arrhenius model and uses experimental data obtained by performing accelerated degradation tests (ADTs) at high temperatures. Using OriginLab, we chose the most suitable model as the degradation model. We calculated the parameter values of the proposed model using the gamma process and Arrhenius model. We estimated the RULs of BJTs using the proposed method.