Research On Robustness Of Urban Public Transport Network Based On Higher-order Network Model

As the main carrier of urban transportation system service function,urban public transport network undertakes the important responsibility of urban residents’ travel and urban transportation.However,the system is affected by many external incidents,such as extreme weather,vicious traffic accidents,etc.These incidents may lead to huge socio-economic losses directly and indirectly.Analyzing and researching the ability of public transport network to resist these incidents is of great significance to public transport network planning and optimization.At present,most of the research on cascading failure and robustness of urban public transport network is focus on a certain transport mode.However,in reality,urban public transport system is a complex system which can be accurately expressed by a multi-relationship complex network which contains multiple transport modes.Therefore,how to build a complex network that can describe the transit system and study the robustness through the complex network has become the focus of this paper.Higher-order network model can introduce different information to expand the traditional complex network model,and forms a higher-order complex network model.Of which,multi-subnet composite complex network model is an important higher-order complex network model.Based on the multi-subnet composite complex network model,this study constructs an urban public transport composite network(UPTCN)containing conventional bus network and subway network.On this basis,considering the travel impedance and edge directivity,an urban public transport composite service network(UPTCSN)is constructed.The topological characteristics of UPTCN and UPTCSN are analyzed by using the public transport data of Beijing and Qingdao.It is found that the UPTCN has the characteristics of small world(WS)and scale-free(BA),the edge weight probability distribution of UPTCSN is concentrated,and the probability distribution of node degree for these two networks is different.The construction of UPTCN provides a basis for subsequent model method research and simulation analysis.Based on information theory,the influence of different relationship strength and interaction degree on material diffusion in UPTCN is studied,and a critical node identification algorithm(IMD)based on mass diffusion is proposed.The IMD algorithm and five other classical algorithms are used to identify the top 20 and top 500 important stations of Qingdao UPTCN respectively.The effectiveness and accuracy of the IMD algorithm are verified by comparing and analyzing the coverage between the calculation results of each two algorithms,comparing with the actual passenger flow and the impact on the network after deleting critical nodes.Critical node identification provides an analysis object for subsequent cascade failure analysis and robustness evaluation.The passenger flow distribution model of node failure in UPTCN is established.Based on this model,the effects of network structure characteristics such as the topology of each subnet,the average degree of subnet nodes,the adjustable parameters of initial passenger load and the importance of nodes on the robustness are studied.It is found that the higher proportion of relationship strength occupied by the subnet topology WS or BA,the greater strength value of the relationship with the lowest average degree,the greater node passenger flow load parameter value,and the smaller edge passenger flow load parameter value,provides the stronger network robustness.The UPTCN robustness of Qingdao and Beijing subjected to random attacks and local attacks is analyzed by simulation.It is found that in a certain range of failure intensity,when the failure times increase,the average travel time decreases,and the reason for this abnormal is explained by OD analysis.At the same time,it is found that random attacks are more likely to cause network failure than local attacks.Further more,the cascade failure model of UPTCN is proposed,and the cascade failure model is transformed into the problem of analyzing network robustness.By using Qingdao UPTCN,the influence laws of external interference strength,tolerance,node and edge coupling strength and important nodes on cascade failure process and robustness are studied.On this basis,the spatio-temporal propagation characteristics of cascading failures in UPTCN and UPTCSN are studied.It is found that cascading failures propagate from multi centers and in the form of a ring,and the distribution of failure nodes presents the form of multiple "peaks".Finally,based on the constructed composite network model,the application research to improve the network robustness is carried out from the perspective of optimization and management.This study puts forward a method to optimize the transfer stations of UPTCN according to the importance of network node complexity,so as to determine the number and location of transfer stations of Qingdao UPTCN.Aiming at the passenger flow at the failed stations,take minimizing the travel time of passenger and minimizing operation cost of conventional bus as objectives,the passenger flow emergency evacuation model and solution algorithm are proposed,which provides the optimal path,train number arrangement and evacuation time of passenger flow evacuation after the failure of Licun subway station in Qingdao.This research constructs a more practical urban public transport composite network,and studies the robustness based on it.The research results can provide theoretical guidance for urban multi-mode public transport network planning and traffic policymaking,and also provide practical countermeasures for network emergency management.