Study on Real-Time Control Based on Dynamic Programming for Plug-In Hybrid Electric Vehicles

This study proposes a real-time control for plug-in hybrid electric vehicles (PHEVs) based on dynamic programming (DP). In order to obtain the optimal controls, DP is first used to solve the driving cycle, and a model-based calibration (MBC) tool is used to generate the optimal maps from the optimal trajectories. Further, a feedback energy management system (FEMS) is developed with the SoC as the feedback variable, which considers the charge and discharge reaction of the battery. To make full use of the energy stored in the battery, combined with the charge depletion-charge sustain (CDCS) strategy, the reference SoC is introduced. Finally, comparative simulation of the proposed real-time controller and DP is performed. The obtained results show that the fuel consumption of the real-time controller is 4.82 L/100 km in the worldwide harmonized light-duty vehicles’ test cycles, which is close to the fuel consumption with DP at 4.69 L/100 km. When the driving distance is known, the SoC can follow the reference SoC to make full use of the energy stored in the battery. The simulation results also indicate that the performance of the controller is good and it can maintain the SoC within the desired range.