A Partitioning Method for Active Distribution Network Based on Hierarchical Spectral Clustering

For active distribution networks with high penetration distributed energy, this paper proposes a partitioning scheme based on hierarchical spectral clustering and the Synchronous Primal - Dual Multiplication Method (SPDMM) optimization. Firstly, Z - score standardization is applied to the distribution network node data to eliminate the impact of different dimensions. Then, an undirected graph is constructed according to the grid topology, and hierarchical spectral clustering is used to cluster the standardized data, dividing the distribution network into several sub - areas. A linear convex optimization model is established for the reactive power regulation problem in each sub - area, and the SPDMM algorithm is used for distributed solution. The SPDMM algorithm fully utilizes the augmented primal - dual Lagrangian function for iterative updates, featuring fast convergence and high flexibility. Simulation experiments are carried out on the improved IEEE - 33 - node active distribution network. The results show that the proposed method can significantly improve the system’s power supply reliability, power quality and dynamic stability.