| A variety of transportation networks,such as roads,railways,and aviation,have a considerable scale.Affected by many factors,different transportation modes in China have been operating independently for a long time,the utilization rate of network capacity is low,and the sharing of resources is poor.“The Outline of the National Comprehensive Transportation Network” proposes to construct a perfect national comprehensive three-dimensional transportation network.Integrate different transportation networks into a multi-level comprehensive transportation network to achieve efficient connection between different transportation modes.In this context,considering the heterogeneity among different traffic networks,this paper describes the comprehensive traffic network as a dependent network,and uses the network theory to analyze the relevant characteristics of the integrated comprehensive traffic network and the optimization of the network structure.First of all,this paper expounds the relevant theoretical basis,analyzes the basic topology indicators of complex networks and the construction methods of dependent networks,and takes the top-level design of national comprehensive traffic network as the guidance,fully considers the correlation and heterogeneity among different traffic networks,and applies the theory of dependent networks to analyze the comprehensive traffic network.Taking aviation and high-speed railway network as an example,the airport and high-speed railway station are regarded as nodes,and the flights and train number are regarded as connecting edges.Based on Space-P network,aviation and high-speed railway networks and aero-high-speed rail comprehensive traffic dependent networks are constructed.the topology structure and network characteristics of the network are analyzed.Aero-high-speed rail comprehensive traffic dependent network and its subnetworks have the characteristics of both scale-free network and small-world network.Secondly,in view of the different physical meanings of connected edges and dependent edges,this paper proposes a load-capacity model of connected edges and dependent edges in the dependent network.By taking into account the characteristics of load distribution on the traffic network when connected edges fail,a load distribution strategy of excess edges in the network conforming to the characteristics of traffic load distribution is proposed.After that,the network attack strategy and robustness evaluation index are defined.The cascading failure model of aero-high-speed rail dependent network is established,and the cascading failure simulation analysis is carried out on MATLAB platform.The simulation analysis shows that attacks on airports and stations are more destructive than attacks on flights and trains.Deliberate attacks,especially dependent edge interface attacks,have a greater degree of damage to the network.After the attack of dependent edge,the dependent connections between networks are interrupted,and the overall efficiency of the network will decrease sharply.The network efficiency is positively correlated with the capacity threshold parameters of the comprehensive traffic network,and the improvement efficiency of the network efficiency decreases with the increase of the capacity threshold parameters.Finally,edge reconnection and edge increase optimization edge adjustment strategies are proposed,and a comprehensive traffic dependent network structure optimization model based on edge adjustment strategies is constructed.The PSO-AIW optimization algorithm is designed,and the optimization analysis of aero-high-speed rail dependent network is carried out on the MATLAB platform.The optimization results show that: The reconnection optimization of the edge with weak load control ability can improve the anti-cascading failure robustness of the network to a certain extent.To some extent,increasing the number of flights or trains can improve the efficiency of the network and enhance the robustness of the three-dimensional transportation network to cope with external risks.Edge enhancement optimization has more obvious effect on the overall network optimization. |