Whale algorithm optimized anode pressure controller for fuel cell systems in ejector recirculation mode

• Employ whale optimization algorithm to mitigate the dependence on experience in fuzzy control logic. • Simultaneous consideration of anode hydrogen pressure fluctuations and purge disturbance. • Fuzzy PI controller optimizations using whale optimization algorithm to enhance system control accuracy. The anode pressure control in proton exchange membrane fuel cells (PEMFCs) significantly influences the stable operation of the hydrogen supply system and the internal gas circulation within the fuel cell. An efficient anode pressure control strategy is imperative for enhancing the overall system efficiency and mitigating lifespan degradation. Effective anode pressure control can prevent hydrogen starvation and instability in output performance under rapid load changes and purge disturbances. Fuzzy control has been extensively employed in anode pressure control studies. However, creating fuzzy rules in the control parameter’s tuning process in existing studies is predominantly dependent on expert knowledge, resulting in concerns about control accuracy. This study investigates the potential of employing the whale optimization algorithm to optimize the selection of fuzzy parameters. We first developed a control-oriented model to address the nonlinearity, coupling, and uncertainty in the hydrogen supply system. Then, based on the model and considering load variations and purge disturbances, we integrated feedforward compensation and fuzzy control into the conventional Proportional-Integral (PI) controller to suppress input disturbances, enhance control accuracy, and reduce the pressure response lag. Finally, an innovative fuzzy PI controller with the whale optimization algorithm is proposed to optimize the fuzzy parameter selection, thereby achieving precise anode pressure control. Simulation tests demonstrate that the whale-optimization-based fuzzy PI control (WFLPIF) reduces a root mean square error by 14.3 % (0.636 vs. 0.742) and a mean absolute percentage error by 28.8 % (0.037 vs. 0.052) compared to conventional PI control, while also outperforming feedforward-compensated fuzzy PI control (FLPIF) by 9.5 % in RMSE and 17.8 % in MAPE. This study substantiates the efficacy of the whale optimization algorithm in addressing the anode pressure stability control challenge of fuel cell hydrogen supply systems.