Scholarlink

Traffic congestion, a major economic and environmental burden, is worsened by inefficient traffic lights leading to higher CO2 emissions. This study explores Deep Q-learning (DQN), a form of reinforcement learning (RL), to optimize traffic light timing and reduce pollution. Using a Barcelona traffic model in SUMO, we demonstrate that DQN-based traffic management can significantly decrease pollution compared to traditional methods. This research highlights the potential of RL for intelligent and sustainable traffic management.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Urban Pollution Reduction via Reinforcement Learning-Based Traffic Light Optimization

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ismail Essamlali,
  • Hasna Nhaila,
  • Mohamed El Khaili

摘要

Traffic congestion, a major economic and environmental burden, is worsened by inefficient traffic lights leading to higher CO2 emissions. This study explores Deep Q-learning (DQN), a form of reinforcement learning (RL), to optimize traffic light timing and reduce pollution. Using a Barcelona traffic model in SUMO, we demonstrate that DQN-based traffic management can significantly decrease pollution compared to traditional methods. This research highlights the potential of RL for intelligent and sustainable traffic management.