Enhancing reactant mass transfer inside fuel cells to improve dynamic performance via intelligent hydrogen pressure control

Abstract

Due to internal mass transfer resistance, the dynamic response of fuel cells often lags the demand power in vehicular powertrains. In this study, an intelligent hydrogen pressure control method based on short-term vehicular power demand prediction is proposed to enhance mass transfer inside diffusion layer and improve dynamic performance. Sensitivity analysis confirmed the positive effect of hydrogen pressure on the dynamic response of fuel cells and revealed the necessity of adaptive switching of hydrogen pressure according to power demand. The short-term power demand is predicted using an iterative learning framework (ILF). Then, the variable-pressure control strategy is presented to realize the intelligent mode switching of hydrogen pressure according to the predicted power demand via a three-way solenoid valve. The effectiveness of the proposed method is validated experimentally. The results show that ILF achieves the best predictability for short-term power demand compared with non-iterative learning framework, thus guaranteeing high decision correctness of the prediction-based control strategy up to 99.9%. The dynamic performance of fuel cell is improved effectively, and its delivered energy increases by 6.98%, such that the battery discharge energy can be reduced by 45.4% in a 35-s test, which confirms the viability of the proposed hydrogen pressure control method.