Use Of Model Predictive Control To Improve The Fuel Efficiency Of Plug-In Hybrid Electric Vehicles

As stated by the Research and Innovative Technology Administration (RITA) the energy consumption in the transportation sector in 2009 accounted for a 28.6% of the total energy consumption in the U.S. With those statistics, the improvement of the vehicles fuel efficiency becomes an imperative requirement, as a key factor to reduce the pressure on the environment, satisfy the stronger governmental regulations and reduce the U.S. dependence on foreign oil. Plug-in Hybrid Electric Vehicles (PHEV) are a viable and efficient alternative to address those issues. Although the efficiency of this vehicle configuration is already superior to a conventional vehicle, further improvements can be achieved by using external information, as real time traffic data, to solve the optimization problem. We propose two approaches based on model predictive control (MPC) techniques as a part of the PHEV energy management strategy (EMS). The first approach utilizes real-time traffic data to solve the optimization problem. The second approach allows adaptation to changes in the predicted velocity profile along the route. In order to study the benefits in fuel efficiency of the proposed approaches, the model of a PHEV and an EMS based on a local optimization approach (ECMS strategy) were integrated in an ad-hoc simulator implemented using MATLAB/Simulink. The model and the implementation of the two approaches as a part of the EMS are presented. The results showed that the use of MPC can contribute with up to 7% improvement in fuel economy with respect to a non-predictive approach.