Research on Multi-Objective Cluster Partitioning Method for Active Distribution Network Suitable for Risk Prevention and Control

A multi-objective cluster partitioning method for distribution networks is proposed to address the risk issues of heavy overload or voltage exceeding limits caused by the large-scale integration of distributed photovoltaics into distribution networks, taking into account the ability of distributed resource regulation. Firstly, consider the impact of self-adjustment ability on cluster coupling relationship and overall control ability when accessing distributed resources, and establish a multidimensional cluster partitioning index that considers the adjustment ability of distributed resources. Then, the K-means clustering algorithm is used to improve the Discrete Particle Swarm Optimization (DPSO) algorithm, transforming cluster partitioning into an optimization solving problem. Finally, the correctness and effectiveness of the method proposed in this paper were verified using an actual power grid as a simulation example.