A Flexible and Comprehensive Multi-Agent Reference Architecture for Local Energy Aggregation Approaches

The rise of decentralised energy generation and storage as well as the growing shift from a centralised to a decentralised energy system is creating new opportunities, but also introduces new challenges for various entities in the electricity sector. For instance, while the increase of decentralised energy resources decreases the dependency on external energy providers, responsibilities for an efficient power grid operation, e.g., grid balancing, are shifted more and more to the distribution network. This shift results in an increasing interest in local aggregation approaches, where a set of entities such as energy storage, consumers, infrastructure provider, etc., work together to achieve a common goal such as consumer cost minimisation, while still operating the power grid in a safe manner. To support the design of a local aggregation approach, which properly coordinates entities in the electricity sector and adheres to each stakeholder's needs, a flexible and comprehensive model has to be established. Therefore, this paper develops a multi-agent reference architecture for the development of local (energy) aggregation approaches by applying the use case modelling concepts of the two top-most layers of the smart grid architecture model, deriving agents from the identified goals and interactions between stakeholders, and embeds these agents in a comprehensive model of a local aggregation approach. The software architecture analysis model is applied to evaluate the developed reference architecture regarding modifiability and extensibility.