Multi-agent based road traffic control optimization

摘要

Transportation has been a crucial aspect of human life for decades. Road traffic congestion is a critical socio-economic issue as it causes high fuel consumption, waste of time, increased environmental pollution, frustration and safety issues. Major reason for increased road traffic congestion is the suboptimal use of available resources such as time and road space due to the static traffic signal plan. Traffic density at multiple adjacent intersections is not effectively considered in current traffic signal control plans. Novel perimeter gating control mechanism based on the dynamic traffic conditions in multiple intersections is proposed using multi-agent systems to improve the throughput of macroscopic traffic networks. Multi-agent systems facilitate the communication and coordination between adjacent intersections. Goal is to attain the emerging effect of minimizing the time loss due to traffic congestion over time for a selected area. Simulation of Urban Mobility (SUMO) traffic simulation suite along with Java Agent Development Environment (JADE) is used to assess the proposed perimeter gating control mechanism. Experimental results show that the proposed method reduces the time loss by 34% when compared with static traffic control method.