A Novel Agent-based Intersection Control Method for Urban Traffic

Intersection control plays very important roles in the urban traffic. However, most intersection control using stationary signal-timing plan. It is very hard to get the optimal signal-timing plan based on the real traffic condition. In this paper, a novel agent-based intersection control method for urban traffic is proposed. All the urban traffic entities are abstracted to agent models. The real traffic conditions of intersection are analyzed by the interaction, coordination and cooperation action among these traffic entities agent models. The optimal signal-timing plan is calculated based on the real traffic conditions. Experimental results show the proposed intersection control method can get better passing rate than the conventional method.