Hierarchical Eco-Driving Control Strategy for Fuel-Cell Hybrid Electric Vehicles Under Multiple Signalized-Intersection Scenarios

Vehicle-to-everything (V2X) information can support energy-efficient vehicle operation in signalized traffic. This paper presents a sequential hierarchical eco-driving strategy for fuel-cell hybrid electric vehicles (FCHEVs) traveling through multiple signalized intersections under prescribed signal-phase and road-speed settings. The upper layer employs an A∗-inspired adaptive heuristic graph search in a discretized time–distance graph. A dimensionless motor-power-demand-based numerical ranking score, evaluated at the average speed of each candidate edge under a zero-acceleration edge approximation, introduces approximate powertrain-load information into node ranking. This score follows the supplied numerical implementation and is neither edge-integrated energy nor equivalent hydrogen cost; therefore, the upper-layer procedure is not claimed to inherit the admissibility, optimality, or bounded-suboptimality guarantees of standard A∗ or weighted A∗. The lower layer uses a twin delayed deep deterministic policy gradient (TD3)-based energy management strategy that penalizes raw equivalent hydrogen cost, SOC deviation, fuel-cell degradation increments, and battery degradation increments. The reported simulations show a lower raw equivalent hydrogen cost than the baseline strategy in the tested scenarios, while terminal-SOC correction reveals scenario-dependent trade-offs among corrected equivalent hydrogen cost, SOC regulation, and model-based component degradation indicators. The results support a practical multi-objective balance under the reported deterministic simulation settings rather than uniform superiority in every individual indicator.