Analysis of the current status of control methods for fuel cell air supply systems

Abstract

Proton exchange membrane fuel cells (PEMFCs) are highly efficient energy-conversion devices for hydrogen utilization and exhibit great potential in sustainable transportation. As a critical subsystem, the air supply system significantly influences stack performance, net power output, and operational reliability, whose control technology remains a major bottleneck hindering PEMFC industrialization. This paper systematically reviews classical, advanced, and special-operating-condition control strategies, evaluating their characteristics, merits, drawbacks, and engineering challenges. Conventional approaches, including PID, MPC, and SMC, are well-established in practice but exhibit weak adaptability to complex scenarios. Advanced algorithms incorporating fuzzy logic, neural networks, and reinforcement learning achieve superior control performance yet suffer from excessive computation and hardware constraints. Control schemes targeting dynamic load, cold start, and faults enhance environmental adaptability but are limited by high energy consumption and slow response. Finally, future directions such as algorithm fusion, lightweight design, digital twin, and standardization are prospected. This review provides theoretical support and engineering guidance for the development of PEMFC air supply control systems.