Introduction  The correct topology information recorded by the power supply department can help the staff monitor the power grid information, analyze the faults, and optimize the operation of the power grid to meet the needs of lean and intelligent management of low-voltage distribution networks. At present, the addition of various new types of electricity-using equipment and users has caused the low-voltage distribution network structure to show a continuous change in characteristics, and the line maintenance cost is greatly increased.

  Method  Therefore, the identification method of low-voltage distribution network attribution relationship based on density clustering was proposed. First, the effective voltage data collected by smart meters were extracted to generate a high-dimensional time-series voltage matrix. Then, the t-distributed Stochastic Neighbor Embedding algorithm (t-SNE) and Density-Based Spatial Clustering of Applications with Noise algorithm (DBSCAN) were applied to cluster the voltage data to achieve identification of low-voltage distribution network attribution relationship. Finally, the actual data of a low-voltage distribution network in Sanya City, Hainan Province were analyzed, and the proposed method is compared with other mainstream topology identification methods.

  Result  The analysis results show that the proposed method can achieve more than 95% of identification accuracy, which is higher than other mainstream topology identification methods.

  Conclusion  The proposed method is effective and advantageous in solving such problems, and can provide reference for practical engineering applications and offer a different research idea in the field of topology identification of low-voltage distribution network.