Multiscale and Multiphysical Models and Simulation for Wind Energy

In the project, simulation calculations for the planning and operation of increasingly larger wind energy installations are being enhanced, among other things, using machine learning methods.

Project manager/project management: ForWind – Center for Wind Energy Research, Carl von Ossietzky Universität Oldenburg

Duration of the project: 07/2022 - 06/2026


Key aspects

  • How can the utilization of wind energy be improved through new mathematical methods?
  • How can the simulation of complex flow systems across various scales be achieved in shorter computational time?
  • What role can machine learning play in the further development of numerical flow simulations?

Numerical fluid flow simulations (CFD) are a crucial tool for the design and operation of increasingly larger wind turbines. Until now, however, such high-resolution flow calculations were only possible for individual cases or were only carried out for sub-components such as airfoil profiles. The wind conditions at wind energy plants are an excellent example of a problem on very many size scales - they range from meso-local weather systems to turbulence at the individual facility. However, CFDs are currently associated with high computational effort, making them complex and expensive.

In the research project MOUSE, scientists are investigating new mathematical methods of flow simulation for their functionality and suitability for use in wind energy. The goal is to enable an improved integration of the various scales with the same or even higher accuracy in shorter computation time. In this project, cross-scale approaches in the area of meso- and microscale simulation for site assessment, wind resource calculation, and plant design are being investigated and expanded. On the one hand, meteorological, aeroelastic, oceanographic, and wave models are coupled in a simulation environment, and on the other hand, the models are supplemented by machine learning methods to accelerate very detailed and thus computationally intensive simulations and to improve their accuracy.

The coupled methods are used in the project to develop novel adaptive wind farm controllers and to test them for an improved description of the load dynamics. In addition, new higher-order simulation approaches for wind energy applications are being explored, promising further optimization of computation time. The new approaches are tested for their practical suitability for the wind energy industry. The results of the research project are intended to accelerate the technical development of wind energy plants in order to be able to monitor them better in real-time during operation.

On October 1, 2023, a new CFD cluster, funded by the Investment and Development Bank of Lower Saxony and the Ministry of Science and Culture of Lower Saxony, was put into operation at the University of Oldenburg. This new high-performance computer was specifically designed for the MOUSE project, as the numerical simulation calculations require an enormous amount of computing power.

Coordination and partners

Coordination: ForWind, Carl von Ossietzky Universität Oldenburg

Dr. Gerald Steinfeld
Tel.: +49 (0)441 / 798-5073
Fax.: +49 (0)441 / 798-5099


Here you can find the list of publications of the Project “FlexiWind”.

Simulating wind flows with AI: Here (Windströmungen simulieren mit KI // Universität Oldenburg ( in this publication you can read how better computer models help to build larger wind turbines and regulate wind farms more efficiently. A new research project led by wind physicist Laura Lukassen is laying the foundations for this - with the help of artificial intelligence. Please note this article is only available in German.