Analytical solution and SOC reference planning for energy and pollutants management of Plug-In Hybrid Electric Vehicles

An important control problem in the automotive powertrain domain is related to the relative use of the two engines of Hybrid Electric Vehicles (HEVs). Aiming to minimize energy and pollutants emissions, this online optimal control problem is highly constrained by the automotive context: very low computing resources, hardly predictable future vehicle dynamic, and need for strong algorithmic robustness. In this framework, the paper proposes two enhancements to the literature reference method. First, an analytical resolution of the optimal control problem is explained, allowing both algorithmic efficiency and simplicity. In the context of Plug-In HEVs, this method needs a battery State of Charge discharge reference, as these vehicles aim to empty their battery in a trip between two charging stations. The literature method aiming to plan this reference at the beginning of the trip fails to conserve optimality in certain heterogeneous trips cases developed in the paper. The second improvement proposed in the paper is then a new method using light cartographic information to deal with problematic cases. This method allows to attain online the optimality of an offline Dynamic Programming algorithm benefiting from a total a priori knowledge of the future powertrain operating points.