Primary and Secondary Collaborative Planning Method of the Multivariate-driven Distribution Network Based on Improved K-means Clustering

In order to solve the problems of poor power supply rate and high average load rate of the distribution network, the development form and path of the driven distribution network based on the quadrant method were adopted, and the improved K-means clustering algorithm was used to cluster and analyze the data of the multivariate-driven distribution network. We have established a multivariate-driven planning model that combines primary grid structure planning of distribution networks with automation terminals and created an intelligent distribution network planning result evaluation index system based on cloud matter element theory. The following results were obtained through experiments: when the number of clustering centers was 5, the clustering index was 0.9, achieving the best effect; After the planning of this method, the maximum increase in the power supply rate of the distribution network was 6.01 %, and the average load rate of the line increased by 39.91 %, with a maximum reduction in operating costs of 10.75%. This method can effectively evaluate the results of distribution network planning.