Multi-objective optimization design and numerical analysis of air-cooled proton exchange membrane fuel cell with bamboo flow field based on neural network model

Choosing an appropriate air-cooled proton exchange membrane fuel cells (PEMFCs) cathode flow field is essential for preventing membrane electrode assembly (MEA) dehydration and improving output performance. This study established a novel bamboo-shaped cathode flow field to promote secondary flow, improve oxygen distribution, and enhance membrane hydration. A multi-channel three-dimensional two-phase air-cooled PEMFC model was established to investigate the coupled effects of flow field geometry on oxygen transport and MEA hydration. Furthermore, optimization design of bamboo flow field is carried out based on neural network model and genetic algorithm. This approach captures the nonlinear relationships between geometric parameters and performance indicators, and efficiently identifies the pareto optimal solutions. The results indicate that the bamboo shaped flow field enhances oxygen distribution uniformity, with an increase of 1.9 % to 2.8 %. In terms of preventing MEA dehydration, the water content inside the membrane increased by 3.4 % with the bamboo-patterned flow field. Compared with traditional parallel flow fields, the bamboo section flow field with a width of 1.0 mm has significant advantages in overall performance, with the net output power density increasing by 10.2 %. The optimized bamboo shaped flow field enhanced the current density by 11.1 % and the net output power density by 12.4 %.