Cluster Partition Method for Large Scale Distributed Photovoltaic in Distribution Network

When solving the problems caused by large scale and small capacity PV in the distribution network, it may be necessary to carry out real-time online large-scale processing, which puts forward new requirements for the ability of the distribution network to process PV. Based on the analysis of correlation characteristics between PV and load in the distribution network, this paper selects appropriate indicators, and the electrical distance characteristics between nodes are also considered, the PV nodes in the distribution network are abstracted into one network. The complex network modularity function is introduced and optimized, and the clustering algorithm is used to partitioning the photovoltaic cluster. The analysis of the examples in two scenarios shows that the clustering result obtained by this method has the highest modularity function value and is closer to the detailed model. In the case of large-scale PV access, it is not necessary to manually set the number of clusters, and the photovoltaic cluster partition can be performed quickly and accurately.