Energy Planning for Unmanned Over-Actuated Road Vehicle

Abstract

Energy planning for unmanned road vehicles (URV)s is an important step for the management of the autonomous driving. This energy planning depends on the model for power consumption estimation related to URVs. Generally URVs are over-actuated, and this property leads to multiple kinematic configurations for driving. Consequently, it adds more constraints and more complexity for energy planning. In this paper, a Neuro-Fuzzy model is developed for power consumption estimation for different driving modes configurations of URV. Furthermore, a dynamic programming algorithm is applied to find the optimal velocity profile, and the optimal configuration mode in each segment of the road for an over-actuated URV called RobuCAR, used for experimental validation.