A Remaining Useful Life Prediction Method for Hydrogen Fuel Cell Based on a Multi-phase Wiener Process-based Degradation Model

As an efficient and environmentally friendly electrochemical energy converter, hydrogen fuel cells are widely used in many areas. Nevertheless, the large-scale commercialization of fuel cells remains hindered by durability concerns. Condition-based maintenance (CBM) strategies, leveraging performance degradation analysis and remaining useful life (RUL) prediction, have shown promise in enhancing cell longevity. However, the individual differences and performance recovery caused by fuel cell shutdown inspection pose significant obstacles to achieving precise RUL estimates. Hence, in the paper, a multi-phase Wiener process-based degradation model was proposed for performance degradation modeling and RUL prediction of hydrogen fuel cells. The conventional stochastic process-based model was extended to a multi-stage degradation expression for describing the voltage recovery effect of hydrogen fuel cells due to planned shutdown. A case study, involving a hydrogen fuel cell, was presented to validate the applicability of the proposed method. The analysis results indicated that the proposed method in the paper significantly improves the accuracy of RUL prediction with small amount of operational data compared to conventional methods, providing valuable engineering applications.