Clustering Methodology Assessment of Multiple Sources and Flexible Loads for Optimal Planning of Active Distribution Network

Abstract

With the rapid increase of renewable generation and electric vehicles (EVs) access to the active distribution network (ADN), the load fluctuation and stability of ADN have been affected. To optimize the configuration of AND and get a reasonable optimal result, it is necessary to cluster load. Furthermore, the detailed operation information of ADN can be obtained by using the cluster analysis method. In this paper, the importance of the clustering method has been analyzed. Additional, classical clustering methods and hybrid clustering methods commonly used in ADN are summarized, and the evaluation indexes including cluster shape, bounds between data sets, and clustering accuracy of clustering results are summarized. Finally, the development prospect of clustering analysis technology in ADN and comments for further clustering technology has been proposed.