A Traffic Based Reference State of Charge Planning Method for Plug-in Hybrid Electric Vehicles

Abstract

An appropriate state of charge (SOC) planning is crucial for improving economics of plug-in hybrid electric vehicles (PHEVs). This paper proposed a novel ensemble learning based reference SOC trajectory variation predictor. It can predict the SOC variation of different road segments based on rough traffic information. On this basis, a framework of multi-objective adaptive equivalent consumption minimum strategy (A-ECMS) is introduced. At the long-term global design layer, the proposed method plans the global reference SOC trajectory based on traffic information. In the real-time control layer, a closed-loop controller is used to update equivalent factor according to the error between current SOC and the reference SOC trajectory. Finally, the proposed method is analyzed and compared with the conventional linearly decreased reference SOC planning method. The simulation results prove that the proposed method improves the accuracy and stability of the planned reference SOC trajectory. Moreover, the total cost of the A-ECMS based on the proposed method is reduced by 2.1 % compared to the A-ECMS based on the linearly decreased planning method, which indicates that the reference SOC trajectory planned by the proposed method can effectively reduce the total cost of PHEV.