A Stochastic Approach for PEMFC material identification

The characterization of fuel-cell materials is crucial for addressing the development of advanced functionalized materials and for fitting fuel-cell models, which are used in performance evaluation and device optimization. This identification still remains challenging when dealing with in situ measurements. The presentation regards a method for dealing with this problem that is based on stochastic optimization. Such techniques are usually applied to specific fuel-cell problems, mostly using semi-empirical models. We present an original formulation that makes use of an accurate zero-dimensional multi-physical model of a PEMFC and of two cooperating stochastic algorithms, particle swarm optimization (PSO) and differential evolution (DE), to extract multiple material parameters from a sufficiently large set of experimental data taken under controlled physical conditions. The method is suitable for application in other fields where fitting of multiphysics nonlinear models is involved.