Balance between energy use and comfort
Developing a self-learning control framework based on Reinforcement Learning, that can autonomously learn and adapt to the occupant behavior in buildings to make a balance between energy use and comfort.
The key objective of the PhD project is to develop a Reinforcement Learning-based control framework for occupant-centric operation of energy systems in buildings. This framework can learn the stochastic behavior of occupants as well as other parameters, such as the stochastic renewable energy potential, to meet the requirements of occupants with minimum energy use.
The self-learning control framework is an alternative to rule-based and model-based methods.