Adaptive Energy Management Strategy for Fuel Cell Hybrid Electric Vehicles Based on Dual-Fuzzy Logic and Extremum-Seeking Control

The energy management strategy (EMS) of fuel cell hybrid electric vehicles (FCHEVs) plays an important role in determining both hydrogen economy and the state-of-charge (SOC) regulation and prevention of over-charge/discharge. Existing fuzzy logic-based EMS typically employs a single fuzzy logic controller with fixed SOC thresholds, which struggles to adapt to complex and varying driving conditions. To address this limitation, this paper proposes a dual-fuzzy logic control strategy comprising a primary fuzzy logic controller (FLC1) and a secondary fuzzy logic controller (FLC2), which operate in coordination to achieve refined power distribution through ΔSOC adjustment, defined as the difference between the state of charge (SOC) of a power battery and its ideal SOC (ISOC). A series of systematic simulations are conducted under different ISOC coefficients (0.45, 0.50, 0.55, 0.60, and 0.65), and the results are compared with those obtained from the switch-fuzzy control strategy. The findings indicate that an ISOC of 0.55 yields the optimal comprehensive performance. Building on this, an extremum-seeking control (ESC) algorithm is introduced to perform online adaptive optimization of the ISOC. A comprehensive cost function that integrates hydrogen consumption, power loss, and SOC deviation is constructed to dynamically adjust the ISOC. Simulation verification is conducted on the MATLAB/Simulink R2024b and AVL CRUISE co-simulation platform under the New European Driving Cycle (NEDC). Results demonstrate that the ESC adaptive strategy achieves hydrogen consumption reductions of 10.0%, 34.6%, and 45.7% under initial SOC conditions of 35%, 75%, and 85%, respectively, compared with the switch-fuzzy control. Furthermore, it delivers an additional 1.21% hydrogen saving over the optimal dual-fuzzy control at an initial SOC of 75%. Notably, the ISOC update in the proposed strategy does not require prior knowledge of future driving cycles, and the optimization itself is performed online. However, the ESC hyperparameters (perturbation amplitude, frequency, filter time constants, integrator gain) and the cost function weights are pre-calibrated offline based on typical operating conditions, and the real-time cost evaluation relies on component efficiency signals derived from the vehicle model, providing a practical and intelligent upgrade pathway for fuzzy logic-based EMS in FCHEV applications.