Development of a fuzzy diagnostic model for polymer electrolyte fuel cells

Polymer Electrolyte Fuel Cells (PEFCs) offer a number of advantages over traditional power generation systems, including high efficiency, high power density, and no local carbon-emissions. However, even the best demonstrator projects suffer in lifetime durability; only surviving up to half the current US Department of Energy (2006) targets. Prognostics and Health Management (PHM) has been identified as a methodology that could be applied to PEFCs to enhance and extend functional lifetime. PHM techniques would be applied through the control systems for the fuel cell; monitoring and managing the operational parameters, and measuring state of health. The approach selected in this investigation is to call upon expert knowledge and understanding of the PEFC functionality; this produces a rule-based fuzzy-logic model. This paper introduces a diagnosticorientated fuzzy-inference model of a PEFC. This combines with existing fuel cell control and monitoring processes, to diagnose a range of commonly documented failure modes.