Semantic interoperability for holonic energy optimization of connected smart homes and distributed energy resources

Recent work has attempted to deliver optimized distributed energy resource management, including the use of demand side management through smart homes. This aims to reduce power transmission losses, increase the generation share of renewable energy sources and create new markets through peak shaving and flexibility markets. Further, this leverages the development of product models at the device, building, and network level within the operational lifecycle stage, beyond the conventional role of BIM between design and construction stages. However, the management of heterogeneous software entities, incompatible data models and domain perspectives, across systems of systems of significant complexity, represent critical barriers to sustainable urban energy solutions and leads to a highly challenging problem space. The presented work describes a systemic approach based on the concept of holonic systems, which exemplify the role of autonomy, belonging, connectivity, diversity and emergence across entities. This reduces the decision complexity of the problem and facilitates the implementation of optimized solutions in real power systems in a scalable and robust manner. Further, the concept of a flexibility market is introduced, whereby smart appliance owners are able to sell load curtailment and deferment to a local aggregator, which interfaces between a small number of homes and a distribution system operator. Artificial intelligence is present at each of the entities in order to express constraints, trade energy and flexibility, and optimize the network management decisions within that entity's scope. Specifically, this paper focuses on enabling interoperability between system entities such as smart homes, local load aggregators, and last mile network operators. This interoperability is achieved through ontological modelling of the domain, based on the existing standards of CIM, OpenADR, and energy@home. The produced ontology utilizes description logic to formalize the concepts, relationships and properties of the domain. A use case is presented of applying the ontology within a multi-agent system, which enables the optimization of day-ahead markets, load balancing, and stochastic renewable generation, and closely aligns with the holonic approach to deliver a holonic multi-agent system. The use case assumes a scenario in line with the emerging energy landscape of a district of domestic prosumers, with a high penetration of micro-generation, energy storage and electric vehicles. Initial results demonstrate interoperability between heterogeneous agents through ontological modelling based on an integration and extension of existing standards, which acts as a proof of concept for the approach.