The design, testing and consulting activities of the project's models, entitled TSO_LoadForecasting - Statistical learning algorithms for load forecasting in TSO, was carried out by CPES researchers Ricardo Bessa, Nuno Fidalgo, Ricardo Andrade, José Paulos and Luís Ribeiro, in partnership with the Energy Institute of Johannes Kepler University Linz, Austria.

“INESC TEC has more than 20 years of expertise and experience in this field of R&D, which are clearly displayed in this project. We focused on several fundamental concepts that we have already studied and tested in industrial environments, namely: automatic and manual feature engineering to extract information from raw time series; combination of multiple statistical and machine learning models; analysis of solution scalability, and model complexity versus accuracy”, said Ricardo Bessa, coordinator of CPES.

A key-element of the project is the feature engineering methodology developed by INESC TEC, and the capacity to deliver high accuracy in special days (e.g. holidays, COVID impact, etc)

“It is important to mention that this project not only demonstrates INESC TEC's internationalisation ability - since the Institute had already developed a project in this field for ENTSO-E -, but also shows that R&D in Portugal can be quite distinctive, attracting other European countries”, emphasised Ricardo Bessa.

The INESC TEC researchers mentioned in this news piece are associated with INESC TEC and UP-FEUP.