Presented at the IFAC World Congress 2026, Busan

How can batteries respond to markets in seconds while protecting their health for years? 🔋
This is the question behind our work presented at the 23rd IFAC World Congress 2026 in Busan, Korea 🇰🇷
Our hierarchical approach separates these two timescales: model predictive control operates on a long timescale, accounting for battery health and future value, while reinforcement learning handles fast market decisions. We will soon upload our arXiv preprint with more details! :)
A special thank you to Rasa Pourjam, whose hard work made this presentation possible. Ehecatl Antonio del Rio Chanona and I co-supervised Rasa’s MSc thesis at The Sargent Centre for Process Systems Engineering, and we are incredibly proud of how far he has taken the project.