Industry Project with EL5-Energo: Applied Results

Industry Project with EL5-Energo: Real Challenges, Real Results

Our master's students recently defended projects focused on developing a mathematical model for short-term generation forecasting at the Kola Wind Farm.

The aim was to develop a mathematical model that could improve short-term forecasts of electricity generation at the Kola Wind Farm using real-world data provided by PJSC EL5-Energo.

Over the semester, the students worked on four tasks:

1. Preprocessing a dataset containing 32 features and 12,000 records.

2. Analyzing machine learning methods for forecasting wind speed and electricity generation.

3. Developing and implementing a mathematical model in software.

4. Assessing the commercial potential of the resulting software.

Throughout the semester, EL5-Energo specialists also delivered lectures on the practical challenges of renewable energy generation in the industry.

The project defense was attended by company experts, including the director of Kola Wind Farm LLC and representatives of EL5-Energo's management.

Our students achieved a forecasting accuracy of 84%, compared with 90% for commercial software — an impressive result after just four months of work.

A proud and joyful moment!