MPC-based power management system for a plug-in hybrid electric vehicle for relaxing battery cycling

Power management strategies affect fuel economy, emission as well as other key parameters such as durability of power-train components. Different off-line and real time optimal control approaches are applied for developing power management strategies while the real-time control seems more attractive in the sense that it can be implemented and directly applied for controlling power flow in a real vehicle. One promising example of this type is the Model Predictive Control (MPC)-based algorithm where a utility function is optimized while system constraints are validated all in real time. MPC-based algorithms have been applied by developing simulation and test bench-based experimental works, but the authors have not seen a report of implementing a MPC-based algorithm in a real vehicle in the literature. In this manuscript, a real-time MPC-based algorithm is developed for implementation in a reference sport class series plug-in hybrid electric vehicle under construction and performance results are compared with engine duty ratio (thermostat) control algorithm. The results show almost identical fuel consumptions in both cases while the relaxed battery cycling is observed with MPC-based strategy which shows the possibility of extended battery life time.