| With the continuous development of traffic detection technology, the traditional detection method is difficult to meet the demand of practical application. In recent years,intelligent video surveillance technology was applied to the traffic information acquisition and processing, solve the problem of traffic congestion, has become a key technology in the intelligent transportation system.Although currently established a relatively complete intelligent transportation system, it is difficult to carry out effective data integration and information sharing between the various sub-control system, only a simple collection and processing of information, and not through the video network sharing and traffic control strategy together, causing the linkage mechanism between the control systems are not formed.This paper is mainly based on National Natural Science Foundation Project(F030209) "Research on Coordinated Operation Linkage Control of Urban Traffic Signal Region Based on Agent and Evolutionary Game", to carry out the key technology of intelligent urban traffic management and control research.The use of video recognition related technology is to achieve real-time extraction of traffic flow parameters, real-time traffic state identification, combined with agent technology,according to the correlation of adjacent intersection traffic flow, through the information interaction between the adjacent intersections to adjust the control cell, and understand the simple degree and emergency degree of coordination task, to take timely measures to control linkage.Firstly, according to the real-time of traffic demand, traffic parameter extraction model is constructed based on virtual coil, describes the implementation process of vehicle detection and vehicle tracking, and analyzes the key elements of parameter extraction, points out the vehicle detection and tracking is the key point of traffic parameters extraction.Seceondly, acccording to the analysis of temporal and spatial variation of traffic parameters. we put forward the quickly identify traffic state and transition model based on fuzzy cognitive map, build the traffic parameters relationship diagram, point out the transition transformation of internal and external influence, analyze the process of transition transformation.Thirdly, according to the analysis of traffic congestion generation and propagation based on the congestion propagation model of system dynamics, determine the key node using the node contraction method, we can get the importance sequence of a key node.According to the laws of diffusion congestion, with the key nodes as the control cell center, the shortest path impedance as constraint conditions, we can determine the scope of the traffic cell. According to the degree of traffic association and similarity of nodes and the associated path to be divided, we put forward the control cell dynamic adjustment and optimization model based on the node contraction method.Fourthly, according to the multi-Agent coordination joint control should not only consider their own operating results, but also consider the impact of their own emissions downstream traffic intersection, each intersection is seen as an Agent, and the establishment of signal optimization model based on genetic algorithms.According to the simple and emergency degrees of traffic assignment, Agent coordinated control selection model based on the theory of fuzzy is set up. According to direct trust coordination and indirect trust coordination to determine the coordination team, build a virtual control cell, the coordination joint control model is proposed based on multiple-Agent.Finally, according to the original system updated and improved. and requirements of interconnection between different functions Agent, information transmission and data sharing, with the thought of integration, we design an urban traffic coordination joint control system based on multi-Agent.Through the key technologies study of video identification and multi-Agent coordination joint control, we can effectively improve the speed and accuracy of the extracted traffic parameters, and enhance the traffic signal control capacity and coverage, to achieve the purpose ofurban traffic in real-time,multi-point and cooperative control.The study in this paper has the guidance and reference to alleviate urban traffic congestion. |
