Clustering Technique for Setting Group Based Protection in Multi-Microgrids System

Multi-Microgrids (MMGs) are developing as one of the feasible options for utilizing locally accessible, cost-effective, and environment friendly electricity provided by distributed energy resources (DERs). One of the most significant aspects in MMG is to ensure adequate coordination among protection devices. The topology of MMGs may alter in a number of ways, including islanded mode, connecting to another autonomous microgrids (MGs) or grid linked mode, etc. Depending on the short circuit current level, the protection devices may have various setting groups for each MMG topology. As a result, using separate relay settings for each topology is not a realistic strategy. Therefore, it requires techniques that cluster similar topologies and form groups to provide adequate protection capable of operating in different groups. This paper proposes a possible configuration of the topological changes for MMG and categorizes them into four and eight groups using the k-means clustering technique. The proposed clustering scheme has been validated on a benchmark MMGs test system. The effectiveness of the clustering techniques has been evaluated based on the topological configuration of microgrids.