Plug-in HEV energy management strategy based on SOC trajectory

Abstract

This paper proposes a predictive control algorithm constrained by the state of charge (SOC) trajectory for the plug-in hybrid electric vehicle (PHEV) hybrid system. Firstly, the hybrid system energy consumption model is linearised piecewise, and the mixed logic dynamics (MLD) model of PHEV with the minimum equivalent fuel consumption as the optimal cost function is established. Secondly, long short-term memory network (LSTM) is used to forecast the future vehicle speed through the historical vehicle speed data. Finally, the SOC trajectory curve is obtained as the constraint condition according to the change of vehicle speed. The optimal motor torque control sequence in the vehicle driving speed prediction horizon is calculated by the model predictive control (MPC) strategy. The simulation results on different standard operating conditions show that the energy consumption of the PHEV is successfully reduced under the constraints of SOC trajectory.