Vulnerability Analysis Of Transport Network Based On Multi-layer Complex Network Theory

With the development of society, transport network plays an increasingly important role in the social life. However, with the growing scale of the transport network, its original transport function is damaged due to some unexpected or random events and the vulnerability of the transport network is also presented. In recent years, with studying and applying the complex network theory deeply, studying the transport network has become more scientific through the perspective of complex network theory. Transport network is a typical complex networks which consists of multiple sites(nodes) and a plurality of sections(edges).To apply the complex network theory, it can exhibit good structural properties of transport network and analyze the vulnerability of transport network.At present domestic and foreign researchers have achieved a lot results in studying the vulnerability of the transport network.There are also some researchers have applied complex network theory to study, but some shortcomings also exists. Most of the researchers only focus on the vulnerability of a sub-network of transport network and they don't pay attention to the association between subnets.This paper constructs a multi-layer complex network model and analyzes the vulnerability of nodes and edges at different attack strategies in multi-layer transport network, and also identifies the key components of transport network. The main content are as follows:First, this paper construts a multi-layer network model. Based on the complex network theory and combining the data of railway and air sub-networks, this paper abstracts the sections and sites into edges and nodes, and builds multi-layer complex network model. This paper also set some concepts and statistical indicators of multi-layer complex networks, and makes multi-layer network efficiency as an important indicator of the relative change in the transport network vulnerability.Second, this paper analyzes the multi-layer network topological properties. After the establishment of a multi-layer model, the average degree of the nodes, average path length, clustering coefficient and efficiency topological properties are analyzed and compared between railway layer, air layer and multi-layer network. By comparing the results of each transport network with the same scale random networks, found that transport network exhibits clustering effect, and air layer network are more effective than railway layer and transport multi-layer network, and transport multi-layer network are between railway layer and air layer in connectivity and efficiency.Third, this paper analyzes the multi-layer network vulnerability. Making use of multi-layer network efficiency indicators, this paper analyzes the vulnerability of nodes and edges under random or systematic attack in railway layer, air layer and multi-layer network. By comparing of the vulnerability of nodes or edges between different networks, we can conclude that multi-layer network was better than sub-network in resisting vulnerability.Fourth, this paper uses the proposed method to identify key components. This method is based on two criteria: first named structure key component is based on the structure of the network topology, and second named performance key component is based on the network performance changes. This paper identifies the key component of single network of railway and air as well as transport multi-layer network. For existing transport network, this paper proposes some protection policy to reduce its vulnerability.