BOARD

NOTICE

HOME > Board > Notice
[Prof. Tariq/usama학생,김상윤학생, Energy/JCR5%] PPO 강화학습 에이젼트기반 multi-agent control under load variations
  • 관리자
  • |
  • 336
  • |
  • 2025-06-11

[Prof. Tariq/usama학생,김상윤학생, Energy/JCR5%] PPO 강화학습 에이젼트기반 multi-agent control under load variations​

 

Good afternoon and good news.

Prof. Tariq at Dongguk Univ  (eX Emselian), Usama Ali (Ph. D student) , Sangyoun Kim(Ph. D student) 's paper has been accepted at Energy journal (JCR Top 5% journal ).
Great paper indeed. Congratulations Prof. Tariq.
Topic : PPO-based Multiagent Reinforcement Learning control under diverse occupancy patterns ( Useful for load or peak variations situation)

작년 연구실 박사졸업하고 3월 동국대 외국인 교원으로 임용된 Tariq교수님 논문이 Energy journal (JCR Top 5% journal )에 게재 승인되었습니다.
박사과정인 Usama Ali학생 , 김상윤학생이 공저자로 참여하였습니다.
연구주제는 PPO 강화학습 에이젼트기반​ 운전 제어 전략으로 특히 일일 변화 또는 주간 변화, 계절별  load변화가 심한경우 유용한 방법). 

연구실 박사후 연구원인 mohammad박사가 서울 사립대 화학공학과 외국인 전임교원으로 임용 예정입니다. 또 다른 좋은 소식 ..

Shahzeb Tariqa+, Usama Alib+, Sangyoun Kimb, ChangKyoo Yoob*, Multi-agent distributed reinforcement learning for energy-efficient thermal comfort control in multi-zone buildings with diverse occupancy patterns,  Energy (ISSN: 0360-5442, SCI, Top 5% journal - THERMODYNAMICS), Elsevier, 332(9), pp. 137082​  (Ack 2개.: 2021R1A2C2007838 and DGU MOE과제) (2025.09)


Keywords : Multizone building; Energy efficient HVAC control; Transfer reinforcement learning; Decentralized control; thermal comfort management.

Research Highlights

Ø  Decentralized multi-agent thermal comfort control in multizone buildings.

Ø  Transfer reinforcement learning enables control policy transfer across buildings.

Ø  Compared to rule-based control, the proposed method reduced grid import by 51%.

Ø  A floor-specific reward structure can further reduce energy consumption by 23%.

Ø  Transfer learning reduced grid imports by 14.4% in limited data scenario.

image.png


image.png
image.png
image.png
image.png
image.png
image.png
image.png
image.png
image.png
image.png
image.png

 

 
이전글 [SI/Applied Soft Computing] Interpretable Reinforcement Learning
다음글 2nd Application for 2025 Fall Semester KHU Graduate International Admission (Jun. 13th 2025 (Fri) 10:00 ~ Jun. 20th 2025 (Fri)
목록