A high-fidelity real-time capable dynamic discretized model of proton exchange membrane fuel cells for the development of control strategies

Abstract

Advanced modeling tools are widely applied to enhance control strategies of proton exchange membrane fuel cells. Existing dynamic fuel cell models, however, still face a conflict between computational speed and simulation accuracy. To overcome this contradiction, a novel discretized fuel cell model is presented based on existing segmented modeling approaches. With a new twofold subdivision method and a newly designed interpolation algorithm, this model allows accurate investigation of the dynamic temperature, two-phase water, and current density distribution. Compared to the finite element method, the simulation duration of hours is reduced to minutes, while the characteristic of high-fidelity is maintained. Model validations are performed on both cell-level and stack-level to demonstrate the model's high accuracy and real-time simulation capability, illustrating the high potential for applying the model in the development of fuel cell control strategies.