On the optimal speed profile for eco-driving on curved roads

The horizontal curvature and a driver's behavior on curved roads significantly affect vehicle fuel economy. This paper describes how optimal speed profiles can minimize fuel consumption for a vehicle travelling on a curve. Previous studies suggested that sustaining a constant speed throughout a level road is the optimal measure for conserving fuel, within certain bounds. Based on the established vehicle dynamic model and an instantaneous fuel consumption model, the optimal constant speeds corresponding to circular curves with different radii can be derived. When entering or departing a curve for additional study, a dynamic programming algorithm is tailored to obtain the optimal speed profile in the vicinity of the curve. The algorithm is verified by using co-simulation of CarSim and Matlab/Simulink, and the results show the algorithm can save approximately 5.46% to 17.64% of fuel compared with the deceleration and acceleration modes of typical driver model controlled by a proportional integral controller. This technology can not only improve the conventional vehicle fuel economy by taking into account the horizontal curvature but also provide decision-making reference for the autonomous vehicle speed control.