基于时间价值和深度强化学习的信号灯控制研究
Study on Signal Control Based on Value of Time and Deep Reinforcement Learning
廖杰 1梁成雄1
作者信息
- 1. 中国科学技术大学软件学院,江苏 215127
- 折叠
摘要
阐述一种基于时间价值和深度强化学习的信号灯控制方法,实现对信号灯路口进行实时控制,并以最小化路口总的浪费时间价值为目标进行控制.通过仿真实验对比传统固定的信号灯控制方法,证明该方法能够大幅减少系统总浪费的时间价值.
Abstract
This paper describes a traffic light control method based on time value and deep reinforcement learning,which is used for real-time control of traffic light intersections and aims to minimize the total wasted time value of intersections.By comparing traditional fixed signal light control methods through simulation experiments,it has been proven that this method can significantly reduce the total time value wasted by the system.
关键词
信号灯控制/时间价值/深度强化学习Key words
traffic light control/value of time/deep reinforcement learning引用本文复制引用
出版年
2024
NETL
NSTL
万方数据