Study On Regional Control Based On Dynamic Flow Of Multiple Intersections

As the continuous improvement of urbanization construction, more and more people join the queues of the car owners. Meanwhile more vehicles swarm into the urban traffic, it greatly increases the complexity of urbanization traffic. Especially in the big city, citizens trip has become a thorny problem. There are many studies about the urban intelligent traffic, and the main algorithms with genetic algorithm, neural network and ant colony algorithm. But as the cost of time and space is too large, it is not the best choice for the high real-time urban traffic.This paper uses the design method of distributed and multi-agent to design the traffic model, which contains five different types of agent of describing different traffic scene entities. Each agent can get traffic data which it needs through communication with its neighbor agents. The paper adopts two different algorithms of single intersection control and global control to a traffic area contains a number of intersections. First, it calculates the signal intersection through artificial metabolic algorithm, and then the Dijkstra algorithm is applied to the global coordinated control of traffic area. Tradition Dijkstra algorithm is used to calculate the shortest path from a vertex to the other vertices in weighted graph, and the paper makes some changes on Dijkstra algorithm.This paper simulates a traffic area which contains9bi-directional four lanes of intersections by using VISSIM(a traffic simulation tool) and Visual Basic platform. VISSIM simulates real traffic scenarios and generates relational traffic parameters after the simulation: vehicle delay, stopped time, the average speed of vehicles, etc. It can evaluate the control method proposed by this paper by the analysis of these traffic parameters. The simulation results show that, the traffic control method in solving the congestion problem of regional transportation appears more effective in different traffic flow.