Research On Structural Characteristics And Anti-attack Capability Of Beijing Subway Based On Complex Network

With the development of urbanization, the traffic pressure of Beijing is increasing.Subway network as an effective way to alleviate the traffic pressure of Beijing is worthstudying. Complex networks theory is applied in this dissertation, combining with thesubway passenger flow data, to study the structure features and the anti-attack capabilityof Beijing metro.This dissertation collected the subway stations and lines from the official websiteof Beijing Subway Bureau, on which the author drew to construct subway networkswith224nodes, in which stations were treated as nodes (the transfer station wascounted once) and the connections between nodes were treated as edges. Suchparameters as degrees, betweenness, and clustering coefficient were utilized to analyzethe features of the subway network. The results show that the connectivity of Beijingsubway network is not high thus the tightness of the connection still has much room forimprovement. So passengers need to take more sites and change several times to reachthe destination. In addition, the spatial distribution of the network is so distinctive thatthe subway administration can enhance the management level on local subway lines orstations based on the existing space clusters division situation.On the basis of the subway stations, a weighted subway networks with directionwas built, in which two nodes were connected with one edge based on the flowdirections of passengers and every edge owns a weight with the number of passengerswithin one certain period. This dissertation also probed into the classification feature ofthe nodes, the key routes distribution, and characteristics of the evolution of passengernetwork during peak periods in a day as well as in a week to understand the weightedsubway networks with direction. The conclusions are as follows: first, the vital nodesplaying roles in the connection are not just limited to the transfer stations in theun-weighted network, but also the nodes with a huge passenger flow. Second, there isno obvious small passenger network community in the weighted network withdirection.Almost all of the stations are closely connected to each other. Therefore, the attention should be paid to the nodes with high traffic flow and the transfer station onthe basis of partition block administration. Third, there is an obvious evolutioncharacteristic of passenger flow, so the metro can allocate the personnel to improve theworking efficiency according to the different passenger distribution.In the end, the author did station and lines breakdowns simulation to investigatethe anti-attack capability of the subway networks. The results show that the maliciousattacks on the subway network probably cause significant security risks. So the averagedistribution of the transfer station should be considered before the construction of thesubway network and keeping eyes on security measures for high traffic nodes.