Abstract
Proton exchange membrane fuel cell (PEMFC) exhibits significant promise in generating power from hydrogen energy. Operating parameters exert a direct influence on both the power output and the uniformity of oxygen distribution within the PEMFC. Therefore, quantifying the impact of operating parameters and identifying the optimal operating conditions are pivotal to enhance the performance and extend the lifespan of the PEMFC. To this end, a two-stage framework leveraging interpretable machine learning and multi-objective optimization is proposed. In the first stage, an interpretable surrogate model for the PEMFC is established. The impacts of single parameter and pairwise parameters on the power output and the oxygen distribution quality are quantified. Moreover, the decision variables are selected for the second stage. In the second stage, the optimal operating parameters are determined via multi-objective optimization. The results from the first stage suggest that operating voltage and pressure have the highest cumulative contribution for both power density and oxygen distribution quality. The influence mechanism of one operating parameter on the relationship between another operating parameter and the research target is clearly quantified. The findings from the second stage indicate that power density increases by 32 %, 27.36 %, and 32.58 % for three optimized solutions, respectively, while the standard deviation of oxygen molar concentration in the selected operating condition is reduced by 29.66 %.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call