DIET Controller: Dynamic Indoor Environment using Deep Reinforcement Learning
Event details
Date | 21.07.2023 |
Hour | 16:00 |
Speaker | Arnab Chatterjee |
Location | Online |
Category | Conferences - Seminars |
Event Language | English |
Arnab Chatterjee (Integrated Comfort Engineering Lab, EPFL) has successfully presented the private defense of his PhD thesis: "DIET Controller: Dynamic Indoor Environment using Deep Reinforcement Learning". This research was conducted under the supervision of Prof. Dolaana Khovalyg, Head of ICE Lab, EPFL. The public defense will take place on Friday, 21 July 2023, at 16:00 at the Smart Living Lab in Fribourg (room HBL0 21A).
Heating, Ventilation, and Air Conditioning (HVAC) Systems utilize much energy, accounting for 40% of total building energy use. The temperatures in buildings are commonly held within narrow limits, leading to higher energy use.
A field study was performed in two office buildings in Switzerland, and the measurements illustrated that the air temperature was relatively steady for most of the hours; it was higher than that prescribed by the building standards in Switzerland. Thus, designing energy-efficient building thermal control policies to reduce HVAC energy use while maintaining a dynamic indoor environment is essential. Also, such an environment might not be the best for long-term exposure for the occupants.
In his PhD thesis, Arnab Chatterjee proposes a deep Reinforcement Learning-based framework for energy optimization and healthy thermal environment control in buildings. As an emerging control technique, deep Reinforcement Learning (DRL) has attracted growing research interest and demonstrated its potential to enhance building performance while addressing some limitations of other advanced control techniques.
Read full abstract
Heating, Ventilation, and Air Conditioning (HVAC) Systems utilize much energy, accounting for 40% of total building energy use. The temperatures in buildings are commonly held within narrow limits, leading to higher energy use.
A field study was performed in two office buildings in Switzerland, and the measurements illustrated that the air temperature was relatively steady for most of the hours; it was higher than that prescribed by the building standards in Switzerland. Thus, designing energy-efficient building thermal control policies to reduce HVAC energy use while maintaining a dynamic indoor environment is essential. Also, such an environment might not be the best for long-term exposure for the occupants.
In his PhD thesis, Arnab Chatterjee proposes a deep Reinforcement Learning-based framework for energy optimization and healthy thermal environment control in buildings. As an emerging control technique, deep Reinforcement Learning (DRL) has attracted growing research interest and demonstrated its potential to enhance building performance while addressing some limitations of other advanced control techniques.
Read full abstract
Links
Practical information
- General public
- Free