Research On Traffic Congestion Identification And Control Based On Complex Network

The route network is the main carrier of air transport and the basis for the efficient operation of air traffic.Based on the complex network theory,a fishbone network model is established,and the statistical characteristics and survivability of the network are analyzed.The calculation method of network traffic capacity and congestion identification method are studied,and the network traffic model is established.The approximate dynamic programming algorithm is used to control the route flow in the route network.The route network modeling and characteristics analyzing.The navigator and report point are used as the route network node,and the route and the route segment between the nodes are used as the edges of the network,then the fish bone oriented weighted route network model is established.Degree,average path length,clustering coefficient,correlation,resistance to destruction and other indicators are studied.The physical meanings,statistical characteristics and invulnerability of each index in the route network are analyzed The empirical research results showed that the fish-bone route network is characterized by a lack of grouping,a negative matching of route networks,and a node with a large degree or weighted value as the central axis node.The damage resistance of the route network at the point of attack is worse than that of the random point attack.Study on identification of route network congestion based on traffic capacity.The calculation method of air traffic capacity was studied,and the calculation method of capacity based on mathematical modeling method and saturated flow rate method was proposed.After a comparative analysis,an efficient and practical saturation flow rate method was determined.Then the definition and calculation method of the route network node and the overall route network capacity are given.Based on this research,the judging indicators of traffic network traffic congestion are proposed,and the congestion nodes in the route network are identified by simulation.Research on route network traffic control based on approximate dynamic programming.A non-linear traffic flow model is established according to the principle of traffic balance.The implementation-dependent heuristic dynamic programming(ADHDP)structure in the Approximate Dynamic Programming(ADP)algorithm is used to control the flow of incoming route in the route network through autonomous learning and multiple iterations.The simulation results show that the approximate dynamic programming method can adapt to the real-time change of traffic volume,smooth traffic flow,traffic congestion is relieved to a certain extent,and provides new ideas for traffic network traffic congestion control.