Modeling of traffic flow of automated vehicles

Abstract

With the development of near term automatic vehicles following concepts such as intelligent cruise control (ICC) and cooperative driving, vehicles will be able to automatically follow each other in the longitudinal direction. The modeling of traffic flow consisting of such vehicles is important for analyzing the effects of vehicle automation on the characteristics of traffic flow and for suggesting macroscopic control strategies to improve efficiency. Such analysis may also suggest ways for modifying the vehicle control characteristics in order to improve the macroscopic behavior of traffic. In this paper, we developed a mesoscopic and macroscopic model that describes the automated traffic-flow dynamics in a single highway lane. The mesoscopic model describes the speed and density continuously in time and space and at the same time retains the microscopic characteristics of traffic flow. The macroscopic model describes the average speed and density at each section of the lane and at each point in tune. Even though the macroscopic model does not retain the microscopic characteristics of the vehicular traffic, computationally it is much simpler than the mesoscopic one. Simulations are used to demonstrate the effectiveness of these models in describing traffic-flow characteristics. The developed models indicate some similarities, but also some fundamental differences with existing traffic-flow models for manually driven vehicles.