Helmholtz AI researchers success at the BigDEAL Energy Forecasting Challenge

The KIT-IAI team with Benedikt Heidrich, Matthias Hertel, and Oliver Neumann took second place out of over 70 participants in the challenge.

After the Global Energy Forecasting Competition (GEFCom) in 2012, 2014, and 2017, Prof. Dr. Tao Hong organized together with Big Data Energy Analytics Laboratory (BigDEAL) a new challenge in 2022. The theme: peak load forecasting, an application of ML methods which aims to help electric operators choose the best efficient mix production for the next day, satisfying the demand. It’s a tool that prevents energy waste and helps with environmental issues by establishing plans for the use of renewable energy.

The competition started in October 2022, with a total of 78 teams formed by 121 contestants who joined from over 25 different countries. After a qualifying match, only 14 teams reached the last stage and competed against each other to win the final match. 

The final consisted of six rounds with the three tasks of predicting the level, timing and shape of the daily peak loads. The IAI team was the only one to win first place in two categories, peak time and peak shape, which led to its second place in the overall ranking.

The team's forecasting method will be published in IET Smart Gridsoon. Stay tuned!