Data-driven Based PEMFC EIS Modeling with Nyquist Plot.

Electrochemical impedance spectroscopy (EIS) Nyquist plot modeling has been attached great importance to fault diagnosis of proton exchange membrane fuel cell (PEMFC) system. This paper applies Gaussian process regression (GPR) and Bayesian optimization (BO) to the problem of building an adaptive EIS Nyquist plot model of PEMFC. The experimental results show that GPR performs better than multivariate polynomial regression and equivalent circuit model (ECM) method for this task when a small number of training samples are available. Therefore, this method can be a suitable approach for online adaption of EIS Nyquist plot model for fault diagnosis application.