Fuzzy logic energy management optimization through genetic algorithm for plug-in fuel cell electric vehicles

This paper presents a novel multi-objective optimization methodology for the energy management system (EMS) of autonomous PFCEVs equipped with a hybrid energy system (HES) comprising a fuel cell (FC), battery, and supercapacitor (SC). A fuzzy logic (FL) controller is implemented to online power distribution, and its performance is enhanced through a two-phase genetic algorithm (GA) optimization. In the first phase, the GA fine-tunes the membership functions (MFs) defining the fuzzy sets, while in the second phase it optimizes the rule weights. The novel SoC planning method, based on quadratic programming (QP), uses vehicle trajectory and GPS data to generate optimal SoC profiles under varying initial battery charge conditions. Simulation results conducted in MATLAB show that the optimized controller improves system performance compared to the non-optimized fuzzy controller, reducing operating costs while preventing deep battery discharge. The strategy maintains the battery state-of-charge above critical thresholds, mitigating battery aging, and achieves an operating cost within 98.4% of the global minimum. Additionally, it slightly outperforms a reference online optimization-based EMS while requiring significantly less computational effort. These findings highlight the potential of integrating real-time oriented FL control with evolutionary optimization techniques for robust and efficient energy management in PFCEVs.