An Intelligent Control Method Based on the Hybrid Algorithm for PEMFC Stack Cathode Air-Feeding and Thermal Control

Abstract

The air-feeding system of a proton exchange membrane fuel cell (PEMFC) delivers oxygen for electrochemical reactions while critically influencing stack power, efficiency, and durability. Compared to hydrogen supply, air management poses greater technical challenges owing to the need for precise dynamic control, composition regulation, and impurity tolerance. Thermal management similarly governs reaction kinetics, water–thermal balance, and material longevity. To address the coupling between these two subsystems, this study proposes hybrid intelligent control architecture. For the highly nonlinear air supply system, a nonlinear enhanced sliding mode controller (ASMC) is developed that achieves finite-time convergence with improved response speed and reduced delay. For thermal management, a PID controller optimized by the RIME (Rime Ice Optimization) algorithm is designed. The coordinated strategy maintains optimal reaction conditions and meets dynamic power demands, thereby ensuring safe, efficient, and sustainable fuel cell operation. Simulation results demonstrate that the proposed ASMC reduces rise time by 85.0% compared to model predictive control and by 70.0% versus standard sliding mode control, with steady-state error 60.0% lower than that of fuzzy logic control. In thermal management, the RIME-optimized PID controller achieves rapid temperature stabilization within the optimal range, requiring 22.0% fewer iterations than the marine predator algorithm. The integrated control architecture effectively decouples air supply and thermal regulation objectives, providing a robust solution for PEMFC system operation.