An Agent Model with Adaptive Weight-based Multi- objective Algorithm for Road-network Congestion Management

This paper proposes an agent model with adaptive weight-based multi-objective algorithm to manage road-network congestion problem. Our focus is to construct a quantitative index series to describe the road-network congestion distribution, and use such indexes as weights in the multi-objective algorithm to shunt vehicles on those congested links. First, a multi-agent system is built, where each agent stands for a vehicle that adapts its route to real-time road-network congestion status by a twoobjective optimization process: the shortest path and the minimal congested degree of the target link. The agent-based approach captures the nonlinear feedback between vehicle routing behaviors and road-network congestion states. Next, a series of quantitative indexes is constructed to describe the congested degree of nodes, and such indexes are used as weights in the twoobjective functions which are employed by the agents for routing decisions and congestion avoidance. In this way, our proposed agent model with adaptive weight-based multi-objective algorithm could achieve congestion distribution evaluation and congestion management at the same time. The simulation results show that our proposed approach has successfully improved those seriously congested links of road-network. Finally, we execute our model on a real traffic map, and the results show that our proposed model reduce the congestion degree of roadnetwork, thus have its significant potentials for the actual traffic congestion evaluation and management. Keywords-adaptive weight; agent model; multi-objective optimization; road-network congestion management