Independent Agents for Urban Traffic Control Problem with Mobile- Agent Coordination

摘要

For a intelligent traffic system (used for controlling traffic lights at intersections) to be useful in practice it needs to be adaptive to outer situation and to be scalable for different urban areas and streets. That is because patterns of the traffic networks and intersections are very diverse and changing in time. In this research scalability and adoptability have been in our center of interest. Other parameter which is very important in controlling traffic lights is prediction of traffic. This prediction would require having the big picture of the flowing traffic and thus centralized kind of Control (which conflict with scalability). Our aim was to design a model that is scalable and adaptive and at the same time has a sort of prediction mechanism. To achieve this we have worked on a model with high abstraction which has three levels of control. Every intersection is controlled by its own traffic situation, its neighbor intersections recommendation and a mobile agent that goes through different intersections in a particular section in the urban area and has a wider view of the traffic.