A Multi-Objective Optimization for Clustering Buildings into Smart Microgrid Communities

Abstract

Smart microgrid communities are small groups of buildings and distributed energy resources (DERs) that collectively work to jointly enhance energy security. As a collective system, their energy performance relies on the energy consumption and generation characteristics of constituent buildings. Yet, in practice, the geographical proximity of buildings is often used as the primary clustering criteria for microgrids. In this study, we proposed a novel multi-objective formulation that expands this criterion to consider the heterogeneity and complimentary nature of building load profiles in the clustering process. Specifically, the proposed model clusters households by (i) minimizing the variation in net-energy across the building cluster, and (ii) minimizing the physical distance between buildings contained in the cluster. The study uniquely leverages the augmented ε-constraint method to efficiently populate and analyze Pareto optimal solutions within the resulting tradeoff space. To illustrate the application of the model, it is applied to cluster households within an imagined neighborhood in Ithaca, New York using the Pecan Street Inc. Dataport database. The results illustrate the ability of the proposed model to efficiently design building clusters that balance net energy and physical distance, allowing for increased utilization of DER resources.