Fleet Speed Profile Optimization for Autonomous and Connected Vehicles

Abstract

Cost optimization is a major concern for autonomous electric vehicles. This optimization problem becomes complicated if a group of vehicles as a fleet move along the road. Optimization based on the leading vehicle to generate the fleet speed profile might not guarantee the overall minimum cost of the fleet. In this paper, an optimization algorithm for a fleet of autonomous electric vehicles is proposed using the total cost of the fleet to generate the optimum speed profile so that the overall cost of the fleet is reduced. Maintaining a safe distance with the adjacent vehicles and safe lane changing on a multilane road depends on the accuracy of decision making based on the data coming from the embedded sensors in the autonomous vehicle. Both of those two cases can be satisfied easily if the vehicles are moving as a group on the same fleet speed where the individual speed of each vehicle can be adjusted based on the relative distance with the leading vehicle. An artificial intelligence (AI) based realistic autonomous electric vehicle modeling considering all the route conditions is provided in this paper, and optimization is done for a fleet of two vehicles where the physical models of the vehicles are different from each other. The proposed optimization algorithm shows a reduction of the total cost for the fleet compared to the optimization done based on only the leading vehicle’s cost.