Thijs Peirelinck receives the Encon Energy Prize 2017 for his research on Artificial Intelligence!


The 7th edition of the Encon Energy Prize brings a new winner! Thijs Peirelinck, former Master student at KU Leuven, and recently employed at EnergyVille, received the prize for his paper 'Using Reinforcement Learning for Residential Demand Response of a Building Model in Modelica’. With this paper he collaborates on a self-learning system for heating installations. Thijs received the prize from Minister of Energy, Bart Tommelein.

For the seventh time a jury had to choose the best master thesis on energy and sustainability out of an ever growing number of candidate students. The jury, composed of experts from industry, education and politics, had 5 finalists to choose from: Dietwig Beckers, Jonathan Verbruggen, Karlijn Overes, Merijn Wuyts and Thijs Peirelinck. The quality of the submissions was again exceptionally high this year.

Nevertheless, the jury unanimously decided the Encon Energy Prize 2017 had to be awarded to Thijs Peirelinck. With his thesis, this Master student in the Science of Artificial Intelligence has collaborated on an intelligent system for electrical heating (by means of a heat pump) in houses. The system not only takes into account the desired temperature, but also the cost of electricity at a certain moment. The system aims to heat when electricity is cheap. Already after 8 days the self-learning system will be cheaper than thermostats currently placed in living rooms. The system makes sure to heat the house, within predetermined limits, a bit more than needed when energy is cheap. This way, the heating will be able to wait a bit longer to incite when electricity is more expensive.

Why buying electricity when it’s cheaper?

The system is currently still a theoretical model. Home owners, in contrast to large enterprises, still pay electricity prices that have been determined for a longer period of time. It is most likely that this is about to change. An important part of our renewable energy, like wind and solar energy, is dependent on the weather. The challenge to always comply to the energy demand is therefore becoming more and more important and will continue to grow once traditional power plants will be substituted by renewable energy sources.

When there is a shortage of energy, power stations should be able to generate extra power. But the other way around, it should also be possible to lower the demand of energy. The easiest way to do this is to make energy expensive when it is scarce and cheap when there is a surplus. This can, however, only be effective when enterprises and households also dispose of systems that can postpone their demand for energy.

“The artificial intelligence model developed by Thijs applies this model on domestic heating systems. However, in the future, it might as well be used for charging electric vehicles or for the use of electric appliances such as dish washer or refrigerators.”: explains Robin Bruninx, president of the jury. “This newly developed system can play an important part in greening our energy provision by making sure the energy demand lowers when energy is scarce. With his master paper, Thijs demonstrated the importance of artificial intelligence in the transition to a renewable energy provision. The jury was highly impressed by the ingenious thinking process behind the paper and the clear-sightedness Thijs demonstrated to develop the system.

This feeling was also shared by Minister of Energy, Bart Tommelein: “Climate change has long been looked upon as a challenge for engineers. The best part about the master paper of Thijs Peirelinck is that it demonstrates other application areas are more and more contributing to the energy transition. With his master paper, Thijs proves that in Flanders we can be a pioneer in new technologies such as artificial intelligence. We can therefore indeed be proud of such an advanced thesis!”.

Students working on a master thesis on energy and sustainability can register till 1 July 2018 for the 8th Encon Energy Prize. More information can be found on: