The energy management system of Diesel Hybrid Electric Vehicles allows the reduction of raw emissions and fuel consumption by including them within the controller goals. With the forthcoming Worldwide harmonized Light vehicles Testing Procedure (WLTP), the consideration of transient pollutant emissions will constitute the only way of e ectively adhering to the emission regulations. Particularly, the use of real time optimal control techniques will provide strategies which are both implementable and achieve the best performance. This thesis focuses on the use of optimal control theory to design causal energy management strategies considering both NOx and soot transient pollutant emissions. An approach based on the minimization of weighted fuel, stationary emissions and a penalty on the engine torque variation is developed and solved using Stochastic Dynamic Programming (SDP) and an extended version of Equivalent Consumption Minimization Strategy (ECMS). The resulting controllers are carefully parametrized and tested over di erent conditions. Both controllers result suitable for real time application, reducing the transient emissions over unknown cycles. The introduced penalty shows a great potential for reducing transient emissions for a low fuel increase, while requiring much lower computational resources than model-based approaches.

Outgoing