A comparative study of clustering algorithms for electricity self-sufficient community extraction

摘要

As a first step of implementing an energy resilient community cluster system, we apply clustering algorithms, including two graph partitioning algorithms and a stochastic optimization algorithm, to electricity self-sufficient community clusters in Yokohama city, Japan. The cluster extraction results are compared in terms of the resulting cost function values and the cluster shapes. The comparison result suggests usefulness of each algorithm in different points of views.