Identification of active distribution network island based on graph theory

In this paper, an island identification method based on graph theory for active distribution network (ADN) is proposed. It aims at identifying the operation mode of the ADN after some switch actions under fault conditions or other emergency situations. Firstly, a node-edge model of the ADN is established in which the switch is seen as the node while the line connecting two switches is seen as the edge. Secondly, an adjacent matrix of switches is formed to describe the connection relationship among switches in the ADN. Thirdly, the connectivity judgment of nodes is conducted according to the adjacency matrix, and then a partition matrix of the ADN will be obtained which is used to describe the partition coverage. Finally, based on the partition matrix and considering the electrical equipments connected with the switches, a partition type matrix is obtained to describe the operation mode of each ADN partition. The partition type matrix can be used to determine whether the partition is in island operation mode. The island identification method based on graph theory provides valuable reference information for the protection and control of ADN, which is important for the safe operation of ADN. This method provides a new idea for the island identification of ADN. The effectiveness and validity of the proposed method are demonstrated by the case studies.