Distributed control system architecture for balancing and stabilizing traffic in the network of multiple autonomous intersections using feedback consensus and route assignment method

Abstract

Abstract Autonomous and intelligent system show a remarkable step in urban traffic management. Autonomous Intersection Management (AIM) is an outstanding example of using an autonomous vehicle and wireless communication technology. The traffic performance of a single AIM system has been proved in many works however, traffic in the network of multiple AIMs is waiting for an implementation. Coordination of traffic between intersections in the network is an important step of managing the overall networked traffic throughput. The authors modeled the traffic network with the multi-agents concept and used the discrete consensus algorithm to coordinate between autonomous agents and implemented the rerouting algorithm in order to distribute the excessive traffics to neighbored intersections with the optimal condition. Our target is to have a balance traffic in each intersection and reaches the equilibrium where the stability has been not compromised. The results show that reaching consensus condition will bring the networked traffic to an equilibrium state where a peak traffic will not be happened. In addition, this method shows that when traffic in a network reached consensus, it will also converge to the Nash equilibrium in the finite time.