Development of a Health Monitoring System for condition assessment of truck fuel cell

Abstract

Proton Exchange Membrane Fuel Cells (PEMFC) are a promising power source for automotive applications due to their high efficiency and zero emissions. However, largescale adoption faces challenges related to durability and lifespan, particularly under dynamic automotive conditions that accelerate degradation. This study proposes a novel method for predicting the Remaining Useful Lifetime (RUL) of PEMFCs under such conditions by integrating multiple operational parameters to improve prediction accuracy, unlike traditional models that rely on a single parameter. A Multiple Linear Regression machine learning algorithm, developed in Python, was used to facilitate predictions. The model was validated using three measurement cycles namely: Constant-Velocity, WLTC, and RDC obtained from a fuel cell powered passenger vehicle. The results show that the model performs well with large, consistent datasets, especially under controlled conditions like Constant-Velocity and WLTC. However, it faced challenges with more variable cycles like RDC, suggesting that further improvement is necessary to handle high variability datasets more effectively.