BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:CityLearn: An OpenAI Gym Framework for Grid-Interactive Buildings 
 - Zoltan Nagy - UT Austin
DTSTART:20230224T130000Z
DTEND:20230224T140000Z
UID:TALK194968@talks.cam.ac.uk
CONTACT:Madeline Lisaius
DESCRIPTION:The decarbonization of buildings presents new challenges for t
 he reliability of the electrical grid as a result of the intermittency of 
 renewable energy sources and increase in grid load brought about by end-us
 e electrification. To restore reliability\, grid-interactive efficient bui
 ldings can provide flexibility services to the grid through demand respons
 e. Residential demand response programs are hindered by the need for manua
 l intervention by customers. To maximize the energy flexibility potential 
 of residential buildings\, an advanced control architecture is needed. Rei
 nforcement learning is well-suited for the control of flexible resources a
 s it is able to adapt to unique building characteristics compared to exper
 t systems. Yet\, factors hindering the adoption of RL in real-world applic
 ations include its large data requirements for training\, control security
  and generalizability. This talk will cover some of our recent work addres
 sing these challenges. We proposed the MERLIN framework and developed a di
 gital twin of a real-world 17-building grid-interactive residential commun
 ity in CityLearn. We show that 1) independent RL-controllers for batteries
  improve building and district level KPIs compared to a reference RBC by t
 ailoring their policies to individual buildings\, 2) despite unique occupa
 nt behaviours\, transferring the RL policy of any one of the buildings to 
 other buildings provides comparable performance while reducing the cost of
  training\, 3) training RL-controllers on limited temporal data that does 
 not capture full seasonality in occupant behaviour has little effect on pe
 rformance. Although\, the zero-net-energy (ZNE) condition of the buildings
  could be maintained or worsened as a result of controlled batteries\, KPI
 s that are typically improved by ZNE condition (electricity price and carb
 on emissions) are further improved when the batteries are managed by an ad
 vanced controller.\n\nBio: Zoltan Nagy is an assistant professor in the De
 partment of Civil\, Architectural\, and Environmental Engineering at The U
 niversity of Texas at Austin\, directing the Intelligent Environments Labo
 ratory. Dr Nagy received MSc and PhD from ETH Zurich. A roboticist turned 
 building engineer\, his research interests are in smart buildings and citi
 es\, renewable energy systems\, control systems for zero emission building
  operation\, and the application of machine learning and artificial intell
 igence for the built environment for a sustainable energy transition. He h
 as received the Outstanding Researcher Award from IBPSA-USA in 2022\, seve
 ral Best Paper awards from the CISBAT conference\, Building & Environment 
 journal\, as well as a Highest Cited Paper award from Applied Energy. He o
 rganizes and chairs the workshop on Reinforcement Learning for energy mana
 gement in buildings and cities (RLEM) at ACM BuildSys.
LOCATION:Seminar time is 1pm BST in Room FW 11\, Willam Gates Hall. Zoom l
 ink: https://cl-cam-ac-uk.zoom.us/j/4361570789?pwd=Nkl2T3ZLaTZwRm05bzRTOUU
 xY3Q4QT09&amp\;from=addon 
END:VEVENT
END:VCALENDAR