INVESTIGATION OF SOME DECENTRALIZED FEEDBACK STRATEGIES FOR ROUTE GUIDANCE IN TRAFFIC NETWORKS

Yannis Pavlis and Markos Papageorgiou

The paper presents a feedback route guidance strategy for complex, meshed traffic networks. Essential components of the strategy are simple, decentralized control laws of the bang-bang, P, or PI types that may be designed based on trial-and-error. Simulation investigations demonstrate the efficiency of the proposed strategy for two example networks under several scenarios of demand and incident conditions. Feedback route guidance, though exclusively based on measurable instantaneous travel times (no predictions, no demand or O-D information are provided), is shown to equalize experienced travel times along any couple of used alternative routes in the network, and to considerably reduce travel delays compared to the no-control case.