Fast Planning of Global SoC Trajectory for Hierarchical Predictive Energy Management in PHEVs

Abstract

This paper proposes a novel hierarchical predictive energy management algorithm with fast global SoC planning speed in PHEV. The algorithm consists of two layers：the upper fast global SoC planning level and the lower short-term vehicle speed predictive and model predictive control (MPC) level. The upper level strategy is driven by traffic data and formulates an optimal global SoC trajectory as the constraints of MPC. In the lower level, the short-term vehicle speed is predicted as the disturbance, and via associating with the obtained global SoC, the optimal energy management strategy can be developed to satisfice the fuel consumption requirements in real time. To evaluate the control strategy, we compare the SoC trajectories obtained from fast global planning (FP) and dynamic programming (DP). Then, the curve is applied to the MPC control level (FP-MPC) and compared with the control effect of the MPC based on the CDCS global SoC (CDCS-MPC). Numerical results illustrate the effectiveness of the fast global SoC planning method and the significant impact via introducing the method into the predictive energy management of PHEV.