A Graph Theory-Based Two-Level Spectral Clustering Approach for Active Power Grouping of DER Integrated Power Distribution System

摘要

This paper focuses on a graph theory-based spectral clustering approach for critical infrastructure management on DER integrated power distribution system. The main goal is to develop a virtual clustering framework for the power distribution system based on active power balancing on lines that can provide stable and reliable clusters. First, the proposed approach is discussed based on the metrics and factors that affect the cluster quality. Then the approach is compared with other clustering methodologies presented in the literature. Finally, the advantages of the proposed architecture for clustering a power distribution system is illustrated using IEEE 123 bus network. It is observed that the proposed architecture is stable and reliable when compared to state-of-the-art approaches and improves the cluster quality by 20 to 50%.