Dual Heuristic Dynamic Programming Based Energy Management Control for Hybrid Electric Vehicles

This paper investigates an adaptive dynamic programming (ADP)-based energy management control strategy for a series-parallel hybrid electric vehicle (HEV). can further minimize the equivalent fuel consumption while satisfying battery level constraints and power demand. Dual heuristic (DHP) is one of basic structures ADP, combining reinforcement learning, (DP) optimization principle, neural network approximation function, which has higher accuracy with slightly more complex structure. In this regard, DHP (EMS) designed by backpropagation (BPNN) as Action two Critic networks approximating policy gradient value function concerning state variable. By comparing existing results such HDP-based rule-based strategies, minimum (ECMS), learning (RL)-based strategy, simulation verify robustness economy adaptability power-split proposed EMS to different driving conditions.