Phase Identification Method Based on Electrical Distance Optics Clustering and Linear Regression Analysis

In order to solve the problem of phase identification errors in low voltage station area, a phase identification method based on electrical distance OPTICS clustering and linear regression analysis is proposed. Firstly, the voltage correlation between nodes was obtained by using Pearson correlation coefficient, and the electrical distance between nodes was calculated. The power balance equation was established for regression analysis of each group of nodes. The phase of each group node was determined. However, the regression analysis could not determine all the phase of each group node. Finally, voltage correlation analysis is used to identify the phase of the grouping that regression analysis cannot determine. The effectiveness of this method in phase recognition is verified by actual power grid analysis.