Integrating Traffic Data and Model Predictive Control to Improve Fuel Economy

This paper presents a method for increasing fuel economy using traffic data and a model predictive controller. Using knowledge of the traffic ahead, a vehicle can react to changes in traffic density or speed before they happen, increasing the efficiency of a trip and providing valuable information to the driver. In particular, the traffic information is used to determine a time-varying velocity envelope that the vehicle must satisfy. Then, a vehicle model is used to compute the vehicle speed profile that minimizes fuel use and satisfies the velocity constraints. Simulation results show the feasibility of the proposed approach on a passenger vehicle with minor hardware modifications required for its implementation.