Implementation and demonstration of an automated energy modeling framework for scalable and adaptable building digital twins based on the SAREF ontology

The building sector is currently seeing rapid digitalization through IoT sensor and smart meter networks as well as the development of more advanced building operation algorithms. In this work, a framework for automated energy model development is developed and assessed as a backbone for effective building digital twin applications. The framework is based on the SAREF ontology and its extensions, SAREF4BLDG and SAREF4SYST to represent the most important devices and components in the building, how they behave, and how they interact. Each component is associated with a data-driven model that is calibrated through operational data collected from the building. The framework is demonstrated and evaluated in a case study building considering a teaching room, where continuous commissioning and scenario testing are investigated as services.