A novel hierarchical predictive energy management strategy for plug-in hybrid electric bus combined with deep reinforcement learning

This paper proposes a novel hierarchical predictive energy management strategy combined with deep reinforcement learning (DRL) for a plug-in hybrid electric bus (PHEB). Firstly, a real-world speed profile is used to train the DDPG algorithm to generate the state of charge (SOC) reference intelligently. Then, a hierarchical model predictive control (MPC) strategy is designed to predict the velocity and allocate energy optimally. At last, the superiority of the proposed strategy is validated under another real-world speed profile. Simulation results indicate that the proposed strategy in this research can reduce the total cost by 10.26% than rule-based strategy.