The wind blows where it pleases. This is also experienced by wind farm operators. To accurately estimate the generation capacity of the plants, relying on local power measurement values and parameters from weather models is not sufficient. AI could improve these predictions, providing accurate forecasts for the next eight hours.

The aim of the Temporal Fusion Transformers (TFT) project is to apply this novel model on wind power forecasts. To this end, a large volume of different data is processed in a model, which then calculates a probabilistic power forecast for each plant. That way, spatio-temporal dependencies that are available for different locations and times are pooled. Wind speed data from one measurement location is incorporated into forecasts for other locations.

The project is interesting to:

Direct marketers, grid operators, utility companies

Project partners

Research Field Energy Informatics, Fraunhofer IEE

Project participants: Jonas Koch, Malte Lehna

Project period

October 2020 – March 2021

Jonas Koch

Project manager

Fraunhofer IEE

+49 (0) 561 7294-1756

Share this Spotlight with your network.

Alsland - Inselnetzerkennung mit Methoden der künstlichen IntelligenzAlsland – Detecting microgrids using artificial intelligence