Teodor Åstrand

My PhD project revolves around the use of reinforcement learning techniques for the optimization of wind farm control, and will leverage reinforcement learning’s ability to learn and adapt to changing conditions. From this, a contribution to the field of intelligent autonomous decision-making within renewable energy is the intended outcome. Specifically focusing on incorporating machine learning methods to optimize wind farms from a holistic point of view, while at the same time contributing to making them a reliable and safe option.

​

Supervisors

Tuhfe Göçmen, Nikolay Dimitrov, Pierre-Elouan Mikael Réthoré

I am from Lund, in southern Sweden, where I studied Mechanical Engineering at the Faculty of Engineering (LTH), Lund University. During my Master’s studies, I specialized in Mechatronics, a cross-disciplinary field that combines principles from mechanical engineering, electrical engineering, control theory, and computer science. My interest for interdisciplinary problems led me to a Master’s thesis where I focused on applying reinforcement learning to robotics tasks. Specifically, I explored how to leverage reinforcement learning for an object isolation task using a collaborative robot arm. From this I developed an interest for reinforcement learning and its potential as a powerful tool.

Motivation

My curiosity for the application of reinforcement learning, along with an interest to engage with the topic on a deeper level has been key motivating factors for me pursuing this PhD. This interest is further strengthened by the opportunity to work at the intersection between renewable energy and AI, an area I find compelling from both a technical and personal perspective. Furthermore, I believe that wind energy is a complex and exciting field that allows for exploration of multiple technologies, enabling interesting collaboration and exciting research.