Together with the Automation Systems Group, opens an external URL in a new window of the Institute of Computer Engineering at TU Wien and the Fraunhofer Institute for Wind Energy Systems (IWES, opens an external URL in a new window), the IET has published the DigiWind-Platform architecture open source as a GitHub repository.
Wind energy systems pose special challenges for digital twins
Digital twins are primarily developed for components or very large systems in the wind energy sector. The goal of the one-year DigiWind research project, funded by the vgbe, opens an external URL in a new window, was to extend the architecture of a digital twin for modular structures and open it up for renewable energies to bridge the gap between industrial practice and the technologies of tomorrow.
Digital twins offer many possibilities to evaluate data during operation, to use it and to profitably feed it back to the real system. Specific to the wind energy industry, however, is that the lifetime of the systems is very long, but the environment and the system itself are constantly changing. This can happen, for example, through changes to the wind turbines, such as the retrofit of aerodynamic improvements, but also through adjustments to the controllers or interventions in the environment, such as the construction of buildings. To map this, modular models are needed, as well as the infrastructure to manage the models, the data, and the simulations.
The DigiWind platform is capable of managing modular simulation models and automatically assembling the simulations. The DigiWind GitHub repository, opens an external URL in a new window contains the infrastructure for linking different simulation models, allowing users to independently create their own services that leverage the basic functionality of the platform. Third-party models can be easily integrated via FMUs.
The successful conclusion of a fruitful research project
With the release of the platform, the DigiWind research project has also come to an end. After one year of intensive work on a model for a digital twin for wind energy systems, which included several online workshop weeks and two research meetings of the project partners in Vienna and Bremerhaven, DigiWind has successfully ended. During the collaboration, some further ideas have been formed, having potential for continued research.
For example, Carlotta Tubeuf (IET) is currently researching, as part of her dissertation and in collaboration with Niklas Requate (Fraunhofer IWES), the possibility of using reinforcement learning algorithms to calculate an optimal operating strategy for wind turbines, in which the service life of the turbine can be extended through strategic derating which leads to a maximization of the overall energy yield.
So it remains to look into the future and ask "What will the future of holistic, integrated control of a wind farm look like using the DigiWind platform?"
Related links: (Original publication, Project website etc.)