Traffic Control Model Based on Granulation Configuration of Traffic Flow

Urban traffic system is a complex stochastic system with characters of non-linear, non-structure and strong-correlation. On the base of traffic flow granulated construction selected by suitable temporal-spacial scales and eliminating or weaking link relationship among granulars, multi-scale resolution approach is a simple problem solution with enough accuracy for urban traffic system. The paper sets up a multi-agent traffic control model based on granulated construction and adopts dynamic parameters matching and harmonizing conflict test of case-based reasoning in order to optimize traffic signal control at intersections. Finally, the paper presents the feasibility of the distributed problem solution with granulated construction by a case of traffic signal control of an intersection in Central Business District of Guangzhou, China.