Deep Reinforcement Learning for Dynamic Traffic Signal Control in Congested Urban Grids

Views: 0
Downloads: 0
DOI https://doi.org/10.55923/aksara.v2i1.1
Dian Susanto 1 , Siti Prasetyo 2 , Maya Nugroho 3 , Agung Wahyudi 4
1, 2, 4Department of Informatics, Universitas Indonesia , 3Faculty of Information Technology, Universitas Diponegoro
1 dian.susanto@ui.ac.id , 2 siti.prasetyo@ui.ac.id , 3 maya.nugroho@undip.ac.id , 4 agung.wahyudi@ui.ac.id

Downloads

Fixed-cycle traffic signal schedules fail during sudden urban congestion surges. We develop a multi-agent deep Q-network algorithm that synchronizes adjacent intersections dynamically based on queue length observations from edge cameras, achieving a 28% reduction in overall transit wait times.
Issue Vol. 2 No. 1 (2023)
Section Articles
Pages 63-75
Published 2023-06-30

Most read articles by the same author(s)

No other articles found by this author.