Multi-agent decision-support system for an urban transportation network

This paper deals with an Agent-based Model of a Decision-Support System for an Urban Transport Network. The present Multi-Agent Decision-Support System (MADSS) assists the regulator in the search of the appropriate regulation decisions in case of a bus traffic disturbance. It is composed of a Supervision Module and a Regulation Module that rely on the information given by the Automatic Vehicle Monitoring system (AVM). The supervision Module represents the temporal axis of the network by the agents of type BUS and the spatial axis by the agents STOP. These agents cooperate in order to detect as soon as possible the disturbances. The regulation Module contains the agents INCIDENT that identify the disturbances, the agents ZONEPERT that analyze them and the agents ZONEREG that generate regulation decisions via an anytime evolutionary approach. Hence, through the cooperative aspects of the agent interactions, the MADSS is able to perform a coherent and preventive regulation process.