Study On Vulnerability Of Urban Rail Transit Network Considering Passenger Flow

Urban rail transit is the main mode of transportation for residents in large cities.With the continuous construction of rail transit,the line network is becoming more and more complex.The local function failure of some key stations or line sections may cause the overall decline of network service level or the failure of areas,which will have a significant impact on Residents’ travel.This characteristic of the network is called "network vulnerability" in the complex network theory.The vulnerability of urban rail transit network is mostly studied from the perspective of network topology.However,passenger flow flows in urban rail transit network,which must be closely related to the vulnerability of the network.How to introduce the passenger flow factor into the vulnerability analysis of urban rail transit network has become a problem that needs to be studied.Based on the complex network theory,this paper establishes a passenger flow weighted rail transit network model and a network vulnerability assessment model,and evaluates the vulnerability of Xi’an rail transit network.The main research contents and results of this paper are as follows:(1)Based on the complex network modeling mechanism,a passenger flow weighted urban rail transit network model is established with the section passenger flow as the edge weight.The sum of the section passenger flow of the sections connected to the stations is defined as the node strength,the congestion cost and transfer cost are weighted together to calculate the shortest path distance and weighted proximity centrality,the network topology is combined with the passenger flow to obtain the weighted aggregation coefficient and PR value,and the Floyd algorithm for the shortest path is improved,which lays a foundation for establishing the vulnerability assessment model below.(2)The vulnerability assessment model of urban rail transit network considering network structure,passenger flow,emergency factors and potential damage risk is established.The model takes the size of the largest traffic subgraph,the loss rate of passenger traffic and the network efficiency as the network performance indicators,and uses the changes of the network performance calculation indicators before and after the network node is attacked as the characterization of the vulnerability of the node.The model designs the attack strategy and attack process including random attack,intentional attack,single site attack,multi site attack and continuous attack.(3)Taking Xi’an rail transit network as an example,its characteristics and vulnerability are calculated and analyzed.The results show that: 1)Xi’an rail transit network has scale-free characteristics,and the weighted passenger flow shows small world characteristics,which indicates that the distribution of passenger flow on the online network enhances the heterogeneity of the network,and also indicates that human economic activities affect the evolution of the rail network.2)The node strength of Xiaozhai station and anyuanmen station is relatively high,indicating that they have great potential risk of damage and high vulnerability.The single station attack shows that the weighted network efficiency of tonghuamen station,Beidajie station and Northwestern Polytechnic University station decreases by 22.27%,16.22%and 14.19% respectively after they fail.These three stations are highly vulnerable;The multi station attack shows that the network is very vulnerable to intentional attacks based on the maximum node strength and the number of intermediaries.3)When the performance of the unauthorized network is reduced by 50%,the site failure rate is about 3% under deliberate attacks and about 9% under random attacks;When the weighted network performance drops by50%,the station failure rate is about 2% under intentional attacks and about 7% under random attacks,indicating that the vulnerability of the network becomes higher after considering the passenger flow,and the vulnerability to intentional attacks is significantly higher than that to random attacks.On this basis,optimization suggestions to reduce the vulnerability of passenger flow weighted network are put forward.