Multi-Objective Cluster Partitioning for Photovoltaic-Rich Distribution Networks

To address issues such as over-voltage and power backflow caused by the high penetration of renewable energy integration into suburban distribution networks, this paper proposes a cluster partitioning method based on multi-indicator collaborative optimization. By integrating modularity, photovoltaic (PV) accommodation capacity, and active-reactive power balance indices, a comprehensive evaluation framework is constructed that balances structural compactness and dynamic functional adaptability. An improved adaptive genetic algorithm is designed to achieve efficient global optimization under complex constraints. Using a typical low-voltage distribution network as a case study, the proposed method is validated for its effectiveness in enhancing PV accommodation efficiency, reducing PV curtailment rates, lowering node voltage violation rates, improving voltage stability, and optimizing power balance. This research provides theoretical support and technical pathways for the stable operation and regulation of distribution networks with high-penetration renewable energy.