Design and Implementation of Predictive Energy Management For Range Extended Electric Vehicles

Hybrid electric vehicles (HEV) provide an acceptable compromise in the transition from fossil fuel based to electric based transportation while introducing additional complexity due to added degrees of freedom. In parallel, the increasing connectivity of modern vehicles contributes to the information available for decision making and controls. In this work, a predictive energy management strategy based around model predictive control (MPC) techniques is implemented and tested on model-in-the-loop (MiL) and hardware-in-the-loop (HiL) environments with models mainly derived with first principles. The components of the implementation, including the derivation of the prediction model, constraints, and selection of prediction horizon, are explained in detail. The factors that affect the prediction, i.e. elevation profile of the road ahead and velocity profile, and their effect to prediction are described. Finally, the differences between the MiL and HiL test results are analyzed.