A Topology Identification Method for Low Voltage Station Area Based on Mixed Integer Programming and Data Mining

Low voltage station topology identification is a key link in achieving transparency in the distribution network. In this paper, a topology identification method for low voltage station areas based on mixed integer programming and data mining is proposed. Firstly, principal component analysis and Z-score are used to realize data preprocessing. Secondly, the CCVP (Correlation coefficient of voltage profiles) value between each user and the master wire of the low voltage station area is calculated to preliminarily judge the household transformer relationship. Finally, the mixed integer programming method based on energy conservation is used to identify and make accurate topology. Simulation examples demonstrate the effectiveness of the proposed identification method.