Multi-agent negotiation approach to simulate and analyze route choice behavior

Dynamic route choice is one of the key theories in the Urban Traffic Flow Guidance Systems (UTFGS). With the purpose of giving consideration to the benefits of individual traveler and network system manager, a new dynamic route choice method, from the point of Distributed Artificial Intelligence (DAI), was presented based on Multi-agent negotiation, regarding drivers in the network, the information center and the traffic system manager as different agents to modeling respectively, and then the route choice negotiation model among agents was given. By using Starlogo, the famous Multi-agent simulation software, three different simulation scenarios: none information without negotiation travel, information without negotiation travel, and information and negotiation travel, were simulated and compared. The results indicate the validity and superiority of the negotiation method in satisfying drivers' travel demands and improving the whole network performance.