Site-specific day-ahead forecasting of wind resources at scale

Power forecasting for renewable energy sources is a vital tool in enabling the green energy transition. We recently re-implemented the Vestas power forecasting system to leverage the latest machine learning algorithms and use cloud technologies for higher reliability and scalability. We will present the challenges and our proposed solutions as well as the benefit of modularized models.

• How to scale machine learning to large datasets and many partitioned models
• How to enable non-coders to configure specialized models
• How to work across the organization when doing ML

————————————————————————————————————————————————————

Hans Harhoff Andersen – Data Specialist | Vestas Wind Systems

Hans Harhoff Andersen has a background in experimental Physics from Aarhus University and has worked for the last 5 years as a data scientist and large-scale data specialist at Vestas. He has worked on improving the way Vestas handles big data and high-performance computing to better enable data science and machine learning.