arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2401.08121cs.LGcs.AIcs.SYeess.SY
收藏

CycLight:基于周期级策略学习交通信号协同控制

CycLight: learning traffic signal cooperation with a cycle-level strategy

Southeast University(东南大学)

Gengyue Han, Xiaohan Liu, Xianyue Peng, Hao Wang, Yu Han

AI总结:

CycLight提出一种基于周期级深度强化学习的网络级自适应交通信号控制方法,利用PDQN同时优化周期长度和绿信比,结合去中心化多智能体协同与注意力机制,在SUMO仿真中验证了其优越性和对传输延迟的鲁棒性。

摘要:

本研究提出了CycLight,一种用于网络级自适应交通信号控制(NATSC)系统的新型周期级深度强化学习(RL)方法。与大多数专注于逐步决策的传统基于强化学习的交通控制器不同,CycLight采用周期级策略,利用参数化深度Q网络(PDQN)算法同时优化周期长度和相位绿信比。这种周期级方法有效降低了频繁数据通信带来的计算负担,同时增强了实际应用中的实用性和安全性。研究构建了一个去中心化框架用于多智能体协同,并引入注意力机制以准确评估周围环境对当前交叉口的影响。CycLight在大型合成交通路网中使用微观交通仿真工具SUMO进行了测试。实验结果不仅证明了CycLight优于其他最先进方法的性能,还展示了其对信息传输延迟的鲁棒性。

英文摘要:

This study introduces CycLight, a novel cycle-level deep reinforcement learning (RL) approach for network-level adaptive traffic signal control (NATSC) systems. Unlike most traditional RL-based traffic controllers that focus on step-by-step decision making, CycLight adopts a cycle-level strategy, optimizing cycle length and splits simultaneously using Parameterized Deep Q-Networks (PDQN) algorithm. This cycle-level approach effectively reduces the computational burden associated with frequent data communication, meanwhile enhancing the practicality and safety of real-world applications. A decentralized framework is formulated for multi-agent cooperation, while attention mechanism is integrated to accurately assess the impact of the surroundings on the current intersection. CycLight is tested in a large synthetic traffic grid using the microscopic traffic simulation tool, SUMO. Experimental results not only demonstrate the superiority of CycLight over other state-of-the-art approaches but also showcase its robustness against information transmission delays.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑