Learn more about Agent.HyGrid
The challenges in the context of the energy transition are manifold and complex. The increasing diversification of individual approaches and the apparent necessity for decentral control and coordination of energy production and consumption make a systematic and standardized approach indispensable, not only for cost reasons. This applies especially to the usage of distributed on-site software systems that shall take care of a growing number of tasks, e.g., cost optimization for consumers groups on the basis of price signals (i.e. demand response).
In the context of smart grid research, the paradigm of software agents is well established. Nevertheless, large-scale applications in real-life systems are not yet foreseeable. Most approaches presented in this context provide solutions for individual problems and tasks, but lack a unified and standardized approach. However, this is required for large-scale applications of agent-based systems in real-life energy grids, in order to guarantee information exchange as well as security and safety on the basis of homogeneous development processes and tools.
Therefore, initially unambiguous definitions and a precise separation of energy agents and their tasks in the context of existing heterogeneous systems are required. Additionally, different energy carriers, like electricity, gas or heat, have to be taken into account. To reach the intended goal of cross-domain interoperability and flexibilization, standardized interfaces, models and development processes are essential.
In the course of the project, the required characteristics as well as the applicability of unified, autonomous and distributed software systems shall be examined in order to be able to embed these systems into existing energy infrastructure. In compliance with existing definitions of the term software agent, we call this approach “energy agent” and define it as follows:
An energy agent is a specialized autonomous software system that represents a single energy conversion system and manages its operation, namely its capability to generate, consume and store energy. The energy agent and the technical system it represents are embedded into one or several energy distribution networks, being able to interact and communicate within these networks as well as with external actors.
The primary goal of this project is the definition of a reference development process for unified, yet customizable energy agents. Analogous to established models from software engineering and automation system engineering (e.g. v-model and rapid control prototyping) the development process for energy agents will be systematically examined and specified within this project. The project milestones are (i) specification and modelling, (ii) implementation, (iii) simulation, (iv) test-bed application and (v) application and maintenance of energy agents in real-life energy grids. In this context, gaps between the temporal requirements of simulations and real systems are to be examined and closed if possible, in order to enable the application of the energy agent approach during all steps of the development process in a coherent manner.
Beyond proprietary formats
Most of today’s automation solutions for grid control lack interoperability, calling for the development of comprehensive standards. However, this endeavor is attached with a high degree of uncertainty and might take several years to accomplish, slowing down the progress of the energy transition significantly. In contrast, a homogenous agent-based system, which can be used in simulations as well as in real-life applications, enables a faster and more sustainable development, which leads to an increased long-term investment security.
The practical applicability of our novel and innovative approach of energy agents will be examined in an intra-disciplinary approach in cooperation with partners from the utility industry.
Unlike most other smart grid-related research projects, that are limited to electricity grids, we look at the energy transition in a holistic manner, taking into consideration not only electricity, but natural gas and heat as well. Based on thermodynamic fundamentals and current communication standards, a general framework and reference data model will be developed, which can be applied to different energy systems and leverages on the energy agent as a basis for internal processes as well as for communication. The resulting possibilities for unified cross-domain interactions between different energy distribution systems will be evaluated in small-scale scenarios with by means of multi-agent simulations, leaving room to initiate subsequent research projects. Topics such as necessary regulation and agent behaviors will be dealt with, but are very likely to be subject of future research projects as well.