Assessment of Structural Vulnerability for Power Grids by Network Performance Based on Complex Networks

Complex networks (CN) have received considerable attention recently since the investigation of small-world and the characterization of scale-free have been discovered in many real networks. Power grids have been widely acknowledged as a typical type of CN. Many works have applied concepts and measurements of CN to analyze the structural vulnerabilities or the mechanism of cascading failure in power grids. However, the theory of CN has developed most from the generic physical perspective which focuses on the common features of all interested networks including internet, WWW, social network, and transportation networks etc. Therefore, the initial research works on CN developed many common concepts and measurements which are supposed to be efficient to any kind of complex networks. Whereas, the common features of CN mostly refer to their non-trivial topological characteristics which do not occur in simple networks. In contrast, the functions and objectives of different networks would be totally different and take many specific characteristics that can not be dealt with by general methodologies. When the theory of CN is gradually applied to some fields involving the specific functions and objectives of networks, such as the protection of networked infrastructures, it is unavoidable to adapt the common concepts and measurements according to the considered scenarios. Specially, when we focus on power grids, we would find that their basic functions and final objectives are very different from many other well-known complex networks which are often the references in traditional research. The basic function of power grids is to satisfy the power supply required by load nodes efficiently and flexibly. Different kinds of vertices, such as generation buses, transmission buses and distribution buses, have different positions and influences to the achievements of the objectives of the networks. The general assumption about the shortest path as the distance between two nodes is no meaningful for power grids where all paths between them should be involved in energy transmission. Furthermore, most of works on CN are based on unweighted and undirected graphs where the obvious physical constraints in power system operation can not be effectively taken into account. The load in cascading failure model of general theory of CN has been defined equally to betweenness. This is far from the real physical loads in power grids and the related capacities of components can not be taken into account. In this paper, based on reconsidering several well-known assumptions of general theory of CN with reference …