Novel topology identification method for distribution network based on graph database

Abstract

In the power system, the distribution network is an important infrastructure for safeguarding people's livelihood and the last link for serving residential users. Its safe and stable operation greatly affects the quality and reliability of users' electricity consumption. With the continuous promotion of the smart grid strategy, the power system measurement technology is constantly improved, which provides new ideas for distribution network topology identification, and also makes it possible to accurately estimate the impedance parameters of distribution lines. In this paper, a topology identification method is proposed based on distribution network data analysis. Firstly, the voltage fluctuation principle of load nodes in the distribution network is introduced, which provides a theoretical basis for the extraction of feature data. Secondly, for the three-phase unbalance phenomenon existing in the medium voltage distribution network, an imputation method for the outlet voltage of distribution transformers is proposed. The distribution network is equivalently modeled according to the degree of similarity between the characteristic data of each distribution transformer. Thirdly, the equivalent network is cut by using the standardized cutting method in the spectral clustering analysis method to get the correspondence between distribution transformers and feeders. And the topology reconstruction method of distribution network based on mutual information and maximum spanning tree algorithm is tested.