With the continuous development of China's distribution system, large-scale interconnection of power grids has become an inevitable trend of development, and distribution network as an important part of the distribution system, its safe and reliable operation is an important guarantee of production and life.  In recent years, power outage accidents and chain failures are frequent, which urges us to further explore the optimization of network characteristics of power system from a scientific point of view.  In the actual operation process, the distribution network topology will change irregularly due to uncertain factors such as the change of contact relationship, equipment maintenance, and supply transfer in the case of failure. Therefore, scientific and effective identification of the distribution network topology is very important for the stable operation of the network.  In this paper, a topology identification model construction method based on voltage correlation coefficient is proposed. According to the principle of electrical distance and voltage correlation of nodes, a physical-probability model of switching state and voltage correlation of distribution network nodes is formed as state variables.

Distribution network;Bayesian model;;PPN model;voltage correlation coefficient