Group coordination for agent-oriented urban traffic management

Abstract

Future cooperative traffic management systems will make use of on-board intelligence and of communication among vehicles and traffic infrastructure. In this demonstration, we present a simulation-based approach of applying multi-agent systems modeling and coordination for dynamic traffic management in urban areas. Traffic participants, modeled as agents, act according to their local goals and preferences under the more global constraints of traffic management. Towards this end, our approach employs decentral coordination and cooperation techniques. We demonstrate (i) a group coordination mechanism allowing groups of vehicles to select their common speed based on a chosen route; and (ii) a group-oriented automated driving method enabling vehicle agents to co-ordinate their speed and lane choices. The demonstration uses the AIMSUN traffic simulation system which has been enhanced to support agent-based simulation. © 2012 Springer-Verlag Berlin Heidelberg.