Diagnosis of membrane chemical degradation for a health management system of polymer electrolyte fuel cells

Diagnostics and health management of fuel cells are key aspects for improvement of reliability and durability. To achieve performance and lifetime targets it is necessary for fuel cell operating conditions to be optimally managed. An improved fuzzy inference system, utilizing multiple high priority health sensors, for diagnostics and health management of polymer electrolyte fuel cells is presented in this paper. Due to membrane chemical degradation having a critical impact on fuel cell health; the investigation focuses on diagnosing this degradation. The fuzzy inference system enables connections between the intricate relationships of fuel cell operating conditions and consequential degradation modes. A database of inference rules for diagnostics is developed through the literature and experimental testing. Experimental testing was conducted on two fuel cells with differing cell areas. Results support the diagnosis of membrane chemical degradation and therefore support the validation of the fuzzy inference system. The approach has shown the capability of providing diagnostics for different fuel cell designs. The diagnostic fuzzy inference system enables proactive decision making as part of an improved health management system to increase availability and lifetime of fuel cells.