Motivation:

In order to achieve climate-neutral energy supply, it is necessary to expand and optimise renewable energy sources. In doing so, the way in which they influence each other should also be taken into account.

The control systems of modern wind turbines currently only consider the respective turbine itself and not how other turbines in the wind farm behave. However, it is precisely in their interaction that the operation of wind turbines can be optimised, loads reduced and thus the service life extended. This can increase the overall efficiency and net yield of wind farms.

​

Objectives and approach:

The KI4Wind project aims to research the use of machine learning (ML) methods to identify specific operating and load situations of WTGs in order to reduce physical loads and optimise the electrical output of WTGs. To this end, sensors from the partner Fibercheck GmbH are first integrated into real wind farms with the support of associated partners in order to supplement existing sensor measurement data sets in a targeted manner. On the basis of these real measurement values and specific synthetic simulation data from the research partner involved, agents based on artificial intelligence (AI) are to be trained to control wind turbines, with Turbit Systems GmbH taking the lead. These agents will then be optimised in the field by the consortium.

​

Innovations and perspectives:

The project offers the potential for optimised wind farm operations compared to existing technologies. The models developed in the project will then be integrated as independent modules into the local operations control of wind turbines or applied in a centralised manner in a cloud computing approach for operational optimisation. In the medium term, the project strengthens the competitiveness of German companies in the wind power sector and thus contributes to climate neutrality in the long term.

​​