At present, topology recognition technology finds it difficult to reflect the influence of power flow characteristics on topology identification. The concept of virtual impedance is introduced. This defines continuous variables with electrical meaning between nodes that are positively correlated with electrical distance as virtual impedance. A virtual impedance-based low-voltage distribution network topology identification method is proposed by analyzing the electrical distance between nodes using existing measurement data in the distribution network. First, a multiple linear regression equation is constructed with virtual impedance between nodes as the dependent variable. Then, using ridge regression, the virtual impedance of the regression equation composed of each single-phase meter and junction meter is calculated to quickly identify faulty electrical equipment with incorrect topology relationships. Finally, a verification model based on derivative dynamic time warping (DDTW) distance is established to reconstruct the correct topology relationships of electrical equipment and achieve the correction of the topology relationships in the low-voltage distribution network. The effectiveness of the proposed method is validated based on actual sample data of low-voltage distribution network substations.