A hierarchical energy management strategy for PHEVs: Optimizing SOC trajectory tracking performance using adaptive initial equivalent factor strategy (AIEFS)

This paper proposes a Hierarchical Energy Management Strategy for PHEVs (HEMS-AIEFS), in which an Adaptive Initial Equivalent Factor Strategy was incorporated to improve real-time State of Charge (SOC) trajectory tracking performance. The HEMS-AIEFS operates on a two-layer structure. The upper layer uses a nodal SOC planning method, where dynamic programming (DP) is first applied offline to generate optimal SOC trajectories based on standard and real-world driving cycles. The cycles and SOC trajectories are then segmented, and relevant data are extracted to train neural network models that predict SOC node trajectories for future journeys. The lower layer implements the Predictive Equivalent Consumption Minimization Strategy (P-ECMS) to track the predicted SOC trajectory, incorporating AIEFS to set the initial equivalent factor (EF0). The results demonstrate that the incorporation of AIEFS significantly improves SOC trajectory tracking accuracy in P-ECMS. Compared to using a driving-distance and initial-SOC map method or a fixed EF0 method to set EF0, AIEFS reduces the mean squared error (MSE) between the actual and optimal SOC trajectories by 36.26 %–91.68 %, and decreases fuel consumption by 0.35 %–7.69 % under WLTC × 2 driving cycle, meanwhile, it demonstrates stronger adaptability across different SOC variation scenarios and driving cycles. Compared to Charge Depleting and Charge Sustaining (CD-CS) strategies, HEMS-AIEFS reduces fuel consumption by 1.84 %–8.47 % under various conditions.