Identification of electrochemical model parameters in PEM fuel cells

Abstract

The target in this paper is to show how Genetic Algorithms apply for parameter identification of different fuel cells. Therefore, two electrochemical models have been fitted for three different fuel cells. The data originates in the current vs. voltage curves (polarization curves) from the published literature. The results seem promising — a real-coded Genetic Algorithm seems to provide with the model parameters that take the properties of the fuel cells into account. The test material is, however, too small to draw more solid conclusions.