Multi – objective Cluster Partition Method for Distribution Network Considering Uncertainties of Distributed Generations and Loads

The intermittence and fluctuation of distributed new energy and load make the cluster partition of the distribution network more difficult. In order to obtain more applicable grid clusters, a multi-objective cluster partition method considering the uncertainty of sources and loads is proposed in this paper. Firstly, the source-load uncertainty model is constructed to obtain the typical scenario of source-load, and the typical scenario is used to characterize the system uncertainty. Secondly, a multi-objective cluster partition model with cluster economic cost, net power complementarity, control dimension and corresponding constraints is constructed. Then, in view of the contradictory nature of each objective, the non-dominated sorting genetic algorithm and fuzzy membership function are used to obtain the optimal cluster results. Through the example simulation of the modified IEEE 33 node and compared with the clustering method without considering uncertainty. The results show that the cluster obtained by this method has the characteristics of low control dimension, reasonable resource allocation and low economic cost.