华南理工大学学报(自然科学版) ›› 2020, Vol. 48 ›› Issue (4): 123-131.doi: 10.12141/j.issn.1000-565X.190413

• 交通运输工程 • 上一篇    下一篇

自动驾驶环境下的路网混合流量日变模型

田晟 许凯 朱泽坤 曾莉莉
  

  1. 华南理工大学 土木与交通学院,广东 广州 510640
  • 收稿日期:2019-07-03 修回日期:2019-08-28 出版日期:2020-04-25 发布日期:2020-04-01
  • 通信作者: 许凯(1993-) ,男,硕士生,主要从事城市交通网络研究。 E-mail:ctxukai@mail.scut.edu.cn
  • 作者简介:田晟(1969-) ,男,博士,副教授,主要从事交通运输工程研究。E-mail:shitian1@scut.edu.cn
  • 基金资助:
    国家留学基金资助项目 ( 201706155003) ; 广东省自然科学基金资助项目 ( 2020A1515010382)

Day-to-Day Evolution Model of Road-Network Mixed Traffic Flow in Autonomous Driving Environment

TIAN Sheng XU Kai ZHU Zekun ZENG Lili   

  1. School of Civil Engineering and Transportation,South China University of Technology,Guangzhou 510640,Guangdong,China
  • Received:2019-07-03 Revised:2019-08-28 Online:2020-04-25 Published:2020-04-01
  • Contact: 许凯(1993-) ,男,硕士生,主要从事城市交通网络研究。 E-mail:ctxukai@mail.scut.edu.cn
  • About author:田晟(1969-) ,男,博士,副教授,主要从事交通运输工程研究。E-mail:shitian1@scut.edu.cn
  • Supported by:
    Supported by the Natural Science Foundation of Guangdong Province ( 2020A1515010382)

摘要:

为研究自动驾驶车辆投放对路网流量演化的影响,建立了包含传统车辆和自动驾驶车辆的混合流量日变模型。分析两类流量不同的演化目标,采用前景理论描述传统车辆出行者的出行行为,以路径前景值最大为目标建立传统车辆流量日变模型,以路径的边际阻抗最小为目标建立自动驾驶流量日变模型,最后通过算例模拟路网混合流量的逐日演化过程。结果表明: 传统车辆与自动驾驶车辆流量达到稳定的时间及演化的趋势显著不同,两者相互影响并构成混合流量演化; 自动驾驶车辆比例越高,系统总时间就越少,不同比例下路径流量演化为不同的稳定状态; 流量转移阈值影响混合流量的稳定条件,且其稳定状态不同于无阈值情况; 自动驾驶车辆会影响短期事件发生前后的流量分布。

关键词:

Abstract: A day-to-day model of hybrid traffic flow including traditional and autonomous vehicles was established, in order to study the influence of autonomous vehicles on the evolution of traffic flow. The different evolutionary ob- jects of the two flows were analyzed,and the prospect theory was used to describe the traveler’s behavior. The day-to-day model of traditional vehicle flow was established with the maximum prospect value of the path,and the day-to-day model of the autonomous vehicle flow was established with the minimum marginal impedance of the path. Finally,an example was used to simulate the day-to-day process of the hybrid traffic flow. The results show that,firstly,the time and evolutionary trend of the traffic flow between the traditional vehicle and the autonomous vehicle are significantly different,and the two interact with each other and constitute the evolution of hybrid flow; secondly,the higher proportion of autonomous vehicle,the less of total system time,and the path flow under differ- ent proportions evolves into different stable states; thirdly,the flow transfer threshold affects the stable condition of the hybrid flow,and its steady state is different from the no-threshold condition; finally,autonomous vehicles can affect the distribution of traffic flow before and after short-term events.

Key words: