Real-time Energy Management for Plug-in Hybrid Electric Vehicle based on Economy Driving Pro System

Abstract

This paper proposes a novel method of real-time energy management for the plug-in hybrid electric vehicle (PHEV) based on model predictive control (MPC) and economy driving pro system (EDPS), which the actual battery state of charge (SOC) is implicated into the real-time energy management instead of the hypothesis that the SOC is used for one cycle at each calculate simulation. Moreover, based on the MPC framework, the deep neural network (DNN) and back propagation (BP) algorithm are adopted to predict the near future velocity, the predict results show the DNNs model has well performance compared with traditional BP algorithm based predictive method. With the support of EDPS, which considering the real-time traffic information and tensor completion algorithm, the SOC trajectory line is added for the actual real-time energy management. According to the simulation, the fuel consumption rate improved by approximately 5.1%(DNN) and 3.8%(BP) compared with the rule-based control strategy.