Distributed and Centralized Approaches for Cooperative Road Traffic Dynamics

Abstract

With the advent of new sensing and communication capabilities, vehicles are now seen as moving sensors which stimulate a renewal of traffic management strategies for both network surveillance and control. Indeed, large deployments of intelligent transport systems and more specifically cooperative systems would provide a dynamic and local-based information network, as equipped vehicles are able to disseminate, receive and forward information with the infrastructure and others vehicles thanks to the C2X framework. This appears to be a promising step forward towards more efficient, safe and environmentally oriented traffic management strategies. Methodological challenges arise when dealing with these new communication features. In this paper, we report a research effort dealing with the integration of these new capabilities into traffic models to design cooperative traffic management strategies and demonstrate their benefit on a simple road section. For that, we consider a distributed approach where vehicles can communicate with each other in a given radio interaction range. A linear stability analysis is performed for a given class of car following models. This shows that a distributed cooperative framework can improve traffic flow stability, which is a key to determine whether traffic is likely to enter in congestion or not. The interest of a distributed approach is highlighted and discussed by comparison with a centralized one, which brings up others ways of controlling the traffic flow.