Portable Model Predictive Controller System Design for Demand Side Management using Semantic Web Technologies

A smart building relies on heterogeneous information systems, with data originating from different sources and represented in various formats. Semantic Web technologies allow connecting these disparate data sources by standardizing metadata with ontologies and enriching information with logical relationships. The latter enables not only shared understanding and interpretability by machines but also querying and reasoning capabilities. Brick and SOSA/SSN are such widely used ontologies in the smart building domain. These ontologies are currently used to semantically describe physical or virtual assets in a building. Building on these ontologies, we propose an ontology to describe a Model Predictive Controller (MPC) system. This MPC is designed as a Demand Side Management strategy for an office micro-grid system. The proposed ontology complements the use of Semantic Web technologies in the smart building. It allows the designed MPC application to seamlessly modify to a different building for both simulation and experimental purposes by communicating with necessary data sources with using the knowledge embedded in the semantic graphs.