Name: Hussain Syed kazmi

Partners

KU Leuven

Promotor / Supervisor

Prof. dr. ir. Johan Driesen

Samenvatting van het onderzoek / Summary of Research

Energy used by flexible loads, such as heating a building or charging a battery, can be shifted in time to reduce the carbon intensity. However, understanding the behaviour of such loads, and how they interact with human users as well as ambient conditions is a complex problem. Just as data-driven methods based on artificial intelligence have made breakthroughs in every domain of life, it is also possible to use them to understand this behaviour and the interactions. A key challenge to achieving this in the real world is that such methods require lots of data to generalize well. Transfer learning overcomes this limitation by translating knowledge across different domains and tasks. This is done in much the same way a human driver can drive across multiple car and road types without additional training. There is, of course, some training on the job. Leveraging these models, we show that energy efficiency of flexible loads can be substantially improved using a large scale real world demonstrator project. The same model can also be used for other applications, including making building occupants better aware of the cost and climate implications of their actions.

Volledige tekst van het doctoraat / full text

Examencommissie / Board of examiners

Prof. dr. ir. Johan Driesen (promotor) 
Prof. dr. ir. Joseph Vandewalle (voorzitter/chairman) 
Prof. dr. ir. Lieve Helsen (secretaris/secretary) 
Prof. dr. ir. Geert Deconinck 
Prof. dr. ir. Johan Suykens 
Prof. dr. Ann Nowé , Vrije Universiteit Brussel 
Prof. dr. Damien Ernst , University of Liège