Frequential identification of a Proton Exchange Membrane Fuel Cell (PEMFC) fractional order model

Abstract

The diagnosis of Proton Exchange Membrane Fuel Cells (PEMFC) is based on reliable models and methods able to follow the system operating state. Impedance spectroscopy is usually used to characterize FC systems and deduce impedance models. Indeed, the deformations of the FC impedance spectrum in the Nyquist plane translate eventual malfunctions or degradations of the system as the stack flooding or drying. This paper deals with the FC characterization using fractional order transfer function impedance model and uses frequency identification methods to identify the model's parameters. The originality of the presented model is that it uses fractional order derivatives, which make it more compact. A fractional order model can be identified either by a method estimating derivation orders and coefficients at the same time or by using a prior knowledge about derivation order and thus estimating only the coefficients of the derivation operators. In this article, parametric identification of PEMFC fractional order model is treated as a non-linear optimization problem where derivation orders are a priori fixed using Taylor approximation of Warburg impedance to deduce an explicit transfer function form.