An improved mathematical method for the identification of fuel cell impedance parameters based on the interval arithmetic

In this paper, an interval arithmetic-based method for parameters identification of a fuel cell equivalent circuit is presented. The fuel cell experimental impedance spectrum, acquired through the electrochemical impedance spectroscopy (EIS), is identified using interval-valued parameters of the fuel cell Fouquet model. The method is based on a branch-and-bound technique, further enhanced using the sensitivity in the segmentation steps. The proposed algorithm is optimised to reduce computational burden and storage memory demand to speed up the computation time and enable onboard applications on fuel cells through embedded system. The result naturally takes into account the uncertainties and noise affecting the impedance measurements. This method easily allows to define a safe operating area for the fuel cell, based on the parameter intervals identified in normal operating conditions. Hence, it is easy to compare the parameter intervals associated to a newly acquired spectrum and check whether they fall out from the safe operating area. This might be considered as a symptom of abnormal operating condition, enabling fuel cell diagnosis, more detailed analyses, and the activation of mitigation and protection strategies.