Multi-time Scale Model Predictive Control Framework for Energy Management of Hybrid Electric Vehicles

In this paper a multi-time scale model predictive control framework is proposed and applied in the efficiency and drivability optimization of hybrid electric vehicles. A multi-layer model predictive control concept simultaneously enables a static optimization over a long prediction horizon and the optimization of the transient system response which leads to better drivability. The proposed control architecture is evaluated on a standard driving cycle and on the example of a parallel hybrid electric vehicle configuration. The obtained simulation results indicate an improved performance of the two layer energy management strategy compared to the case when a single layer model predictive control scheme is applied to optimize the fuel economy of a hybrid electric vehicle. Although the concept has been proven on the example of parallel hybrid electric vehicle it holds in general for any other hybrid configuration as well.