<p>This study presents an optimal structure approach for the flow channel to enhance the performance of a proton exchange membrane fuel cell (PEMFC). An integrated methodology combining computational fluid dynamics (CFD), artificial neural networks (ANN), and genetic algorithm (GA) based multi-objective optimization (MOO) was developed to optimize the baffle structure within the flow channel. Three parameters, i.e., baffle height, length, and number, were investigated through CFD simulations to determine their influence on the PEMFC performance, including net power density and oxygen distribution uniformity. To overcome the computational expense and time constraints inherent to CFD, two ANN-based surrogate models were developed using the simulation results to rapidly and accurately predict power density and oxygen uniformity, respectively. The GA-based MOO was subsequently applied to the Pareto Optimal results from the surrogate models to identify the optimal baffle structure that maximizes both net power and oxygen uniformity. The optimal baffle structure was again validated through the CFD simulation. The validation result demonstrated the agreement between the CFD simulation and the ANN-based surrogate model. The MOO case achieved significant improvements in performance metrics, with 6.8% and 5.6% increases in net power density compared to the reference and Level 1 structures, respectively. However, it exhibits a 0.85% lower net power density than the Level 4 case. Detailed analysis revealed that the increasing baffle structure volume enhances flow acceleration and establishes substantial pressure drops, resulting in reduced liquid saturation at the cathode GDL/CL interface and thus improved reactant delivery and water management capabilities. In terms of oxygen uniformity, the Level 4 case shows values comparable to the reference, whereas the MOO case attains the highest uniformity index. Consequently, considering both net power density and O₂ uniformity, the MOO case represents the most appropriate solution. This study provides valuable insights into the complex transport phenomena within PEMFCs and efficient structure approaches for their flow channels.</p>