In the complex network system, the power grid has its own physical and operational characteristics: 1) the branch information of the power grid topology model is not small, but is simple. The node information is not the vertex of translation 1 according to different functions. Its network load has the actual physical properties, and it can be classified without distinction 2) in the power grid, the electric energy can only be saved by the generator. Other nodes can only be used as the gathering point of electric energy and can not generate new energy, which is different from the generation and transmission process of information in general network. In view of the above problems, many scholars have explored the construction of power grid topology model based on complex network theory. The weighted idea is introduced in the literature, and the weighted betweenness of nodes and lines is defined as the vulnerability index of the system, which has a good identification effect for the nodes and lines in the special position of the power grid structure which have a significant impact on the system vulnerability. The paper further studies the relationship between node betweenness and its capacity, and points out that the failure of a few high betweenness nodes will lead to a chain reaction when the capacity is sufficient. Therefore, in addition to comprehensively improving the betweenness capacity of components, improving the structure of the network itself for high betweenness components is more practical to suppress cascading failures. Generally, the weighted dielectric model is used in the above literature, and the vulnerability assessment index of junction is constructed by equivalent fitting the characteristics of components with different voltage levels and load levels. However, an important problem is noticed: in the process of power transmission, the power between buses does not only flow according to the shortest path of the network, and the role of other possible transmission paths should not be ignored in the calculation of network dielectric. In reference 7.29, an improved method is proposed to remedy this defect by introducing the concept of "electrical dielectric" based on the law.