Engineering-oriented modeling and real-time fuzzy PID control of a PEMFC air supply system using AMESim–Simulink co-simulation

The air supply system is a critical subsystem determining the dynamic performance, energy efficiency, and durability of automotive proton exchange membrane fuel cell (PEMFC) systems. Conventional fixed-parameter control strategies and high-fidelity modeling approaches often fail to balance modeling accuracy, control performance, and real-time implementation feasibility for mass production applications. To address this practical engineering gap, an engineering-oriented modeling and real-time fuzzy PID control framework for PEMFC air supply systems is proposed based on AMESim–Simulink co-simulation. A “selective fidelity” component-level model is developed in AMESim, which retains only control-relevant dynamic characteristics while simplifying secondary physical processes, reducing computational complexity by 75% while maintaining dynamic prediction error below 4.2%. A low-complexity fuzzy PID controller is designed and implemented in Simulink, featuring a mass production-calibrated rule base, strict parameter adjustment limits, and lookup table-based inference implementation, with a total execution time of 0.32 ms/control cycle compatible with low-cost automotive microcontroller units (MCUs). The effectiveness of the proposed framework is systematically validated through three-level verification: comparative simulation under step load and Worldwide harmonized Light vehicles Test Cycle (WLTC) driving cycle conditions, bench testing, and real vehicle experimental validation. Simulation results show that the proposed fuzzy PID control strategy reduces air flow overshoot by 58% and settling time by 50% compared with conventional PID control, with no increase in parasitic energy consumption. Real vehicle testing on an 80 kW production fuel cell platform demonstrates that the mean absolute percentage errors for air flow rate and cathode pressure remain below 4.65% and 4.34%, respectively, consistent with simulation predictions. The proposed framework achieves a balance between performance improvement and implementation simplicity, providing a practical and deployable solution for real-time air supply control in mass-produced automotive PEMFC systems.