Research On Complexity Of Expressway Network Based On Traffic Flow State

As a channel connecting different cities,highways not only facilitate people's travel,but also promote economic connections between regions.What's more,as an important part of national economic development,highways activate the economic development of the areas along them.Economic development determines the scale of the expressway network.The development of the expressway network promotes the rapid improvement of the economy.They complement and promote each other.Research on the operation of the highways will help to improve the current state of traffic flow,achieve the purpose of improving economic development and facilitating people to travel.This paper innovatively improves the breadth-first search algorithm according to the situation of highways.It uses this algorithm to search all possible routes and uses toll data to restore every information of driving route.It analyzes the complete driving process and estimate the traffic flow status of each section of the expressway network.In addition,the income of each toll record is calculated according to the actual payment amount of each route passing,and finally the income status of all highways is summarized.It innovatively proposes a complex network analysis method for the expressway network: according to the traffic flow status and profit status of the road segment,establish networks about the traffic flow status and the profit of highways,and analyze the degree and betweenness centrality of the network topology to select the key nodes.The graph neural network algorithm is applied to the traffic flow state network,classifying the network nodes.Analyze the classification results of the key nodes,and find the reasons why the key nodes belong to different groups.This paper uses toll data of some highways from Guangdong Province for case analysis.Using python and pajek as tools,the above method to obtain the highway status,and show it by a heatmap.The traffic flow state is used as the edge weight to the traffic flow state weighted network,and the revenue state is used as the point weight to the revenue state network.Among the nodes whose road network degree value is greater than 2,and the condition that the ratio of the weighted network to the road network degree value is close to 3,7 key nodes are found,whose performed bad actually.,4 traffic flow transit highways and nodes are found according to the betweenness centrality of nodes greater than 0.2,which is a new way to find the expressway transit section in a quantitative way.According to the analysis of graph neural network,the classification of key nodes is different,which is different from the expected effect.It is concluded that the use of graph neural network requires a suitable network construction method.