Prognostics of PEM fuel cells under a combined heat and power profileÕ Õ

Prognostics have started to be applied to Proton Exchange Membrane Fuel Cells (PEMFC). Indeed, it seems an interesting solution to help taking actions that will extend their lifetime. PEMFC are promising solution for combined heat and power generation (μ,CHP). As power suppliers, they cannot afford running to failure. This work presents a prognostics application on a PEMFC following a μ,CHP profile. A critical issue with such a mission profile is to be able to model the variation of the power demand. So a key point of this work is the presentation of a model introducing the time dependency of the mission profile as well as the degradations of different inner components of the PEMFC. This model starts from a classical polarization expression transformed based on a detailed understanding of the degradation phenomena and the introduction of time-varying parameters. This model is able to follow accurately the behavior of the PEMFC during its functioning. It is then used to perform prognostics and predict the future behavior of the stack with a particle filter-based framework. The results are very encouraging as the behavior predictions are accurate, with a low uncertainty and an horizon as great as thirty days.