A GRAPH BASED FRAMWORK TO SUPPORT DATA-DRIVEN URBAN BUILDING ENERGY SIMULATIONS

Urban planners and energy policymakers increasingly focus on sustainable urban development and the challenges of analyzing complex urban energy systems. Current models often lack the integration of diverse urban datasets and do not adequately address the dynamic nature of urban energy demands. This study proposes a data-driven framework that involves data collection and preprocessing, building archetypes, machine learning modeling, and parametric simulation. The novel contribution of this research lies in defining the scope, processes, information, data, and relationships for the ontology of urban building energy modeling, employing a graph-based approach for complex data integration. The proposed methodology is tested in residential buildings in Dublin City to examine and compare the modeling results. The study concludes that the proposed model offers a more comprehensive and adaptable approach to urban energy analysis compared to traditional methods. Furthermore, the study helps stakeholders by providing a scalable and flexible modeling framework for urban energy analysis.