Weibull-based reliability integrated degradation framework in proton exchange membrane fuel cell system

In view of their high energy conversion efficiency, Proton Exchange Membrane Fuel Cells (PEMFCs) are a promising sustainable energy technology with low emissions and rapid start-up capability. However, their long-term durability and reliability remain significant challenges that limit widespread deployment. Accurate lifetime prediction is therefore essential for effective operation and maintenance planning, as performance degradation increases failure risk over time. This study proposes a Weibull-based framework for degradation modelling and reliability analysis of PEMFC systems. The approach estimates key reliability metrics, including failure probability and reliability function, while characterizing aging behaviour through statistical modelling. The degradation characteristics of PEMFC systems are analysed using the proposed Weibull model and compared with exponential degradation modelling, ARIMA and Support Vector Regression (SVR). The Weibull model demonstrated more interpretable reliability-oriented model performance, achieving an R² of 0.9754 and RMSE of 0.0310, compared to the baseline model R² = 0.9692 and RMSE = 0.0347, highlighting its improved capability in capturing nonlinear degradation behaviour. The estimated Weibull shape parameter further indicates evolving degradation dynamics and provides insight into changes in failure rate over time. Through integration of degradation modelling with probabilistic lifetime analysis, the proposed framework enables improved reliability assessment and supports predictive maintenance strategies for PEMFC systems. The results demonstrate that the Weibull-based approach provides a robust and effective tool for linking degradation behaviour with lifetime prediction, contributing to enhanced reliability evaluation in fuel cell applications.