INESC TEC researcher awarded for work on the use of AI in the energy sector

Artificial Intelligence (AI) tools have proven their efficiency in different applications in the energy sector, from forecasting to optimisation and decision-making. This is precisely the starting point of the work developed by Konstantinos Parginos, researcher and PhD student at INESC TEC and Mines Paris Tech, under Marie Curie Artificial Intelligence for the Sciences (AI4theSciences) doctoral program; his research work earned him the Best Poster Award at the WindEurope 2023 Annual Event.

“Optimizing wind energy trading decisions  using interpretable AI-based tools – the symbolic regression approach” is the title of the research on interpretable AI for renewable energy trading – a work that has been acknowledged among more than 250 posters. As a case study, the research focuses on trading the electrical energy produced by a wind power plant, testing different market conditions and scenarios. The goal is to use interpretable AI tools to extract useful and efficient data-based information and enable advanced human-in-the-loop trading strategies.

“We have developed an evolution-based algorithm that includes feedback from human experts and symbolic representation of the trading strategies. The result of using this algorithm is a substantial decrease in the costs of penalising for imbalance in all tested scenarios and markets, corresponding to a reduction of up to 45%”, mentioned the researcher.

According to Konstantinos Parginos, the interpretable AI tools developed should not only perform well in standard situations, but also be able to ensure robust solutions in the event of extreme events.

Professionally, the award represents a validation of the potential impact of this research and a contribution to the advancement and recognition of these tools applied to the renewable energy industry.

“I’d like to thank EAWE for their support and recognition, which inspire me to advance the research, towards new ways to contribute to the development of interpretable AI tools for renewable energy. I am particularly grateful for the guidance and support provided by my co-supervisors, Ricardo Bessa (INESC TEC), George Kariniotakis and Simon Camal (Mines ParisTech). Their advice and experience have been crucial to my research and personal improvement”, explained Konstantinos.

The 2023 edition of Wind Europe, held in Copenhagen (Denmark), was the largest so far, welcoming more than 15.000 participants from all over the world.