Online Energy Management of FCHEVs based on Multi-Time Scale Dynamic Game Theory

Efficient energy management is critical for fuel cell hybrid electric vehicles (FCHEVs). To address non-stationary demand and competing objectives, a multi-Time scale energy management strategy based on dynamic game theory is proposed. The outer layer optimizes hydrogen consumption, high-efficiency operation, and overall economy over a long horizon. The inner layer performs reference tracking and suppresses power fluctuation on a fast horizon, thus improving durability. On this basis, a dynamic-game algorithm with an online multi-objective equilibrium updates weights via constrained projection and smoothed best response. Real-Time trade-offs among hydrogen consumption, efficiency, and fluctuation are achieved. Validation on 10-kW sightseeing cars under typical cycles shows 75.3% operation in the high-efficiency region while meeting constraints on hydrogen use and fuel-cell fluctuation. Average computation time is 38 ms per step, confirming real-Time feasibility.