摘要
城市道路多信号交叉口影响下的行程时间分布及可靠性是交通流理论研究中的重要方向之一。本文基于灰色关联理论建立信号协调控制下的多信号交叉口行程时间影响因素模型。首先,对车牌识别数据进行预处理,得到路段和干线行程时间数据;然后,利用Burr分布和高斯混合模型对数据进行分布拟合,并进行拟合优度检验;最后,利用灰色关联法分析交叉口数量、干线长度和干线流量与行程时间特性之间的关联关系。结果表明,单路段行程时间分布具有明显的双峰现象,高斯混合模型适用于单路段行程时间的拟合;而Burr分布可以较好地描述多信号交叉口干线行程时间分布右偏和高峰值的特征。交叉口数量、干线长度和干线流量与行程时间特性之间有较强的相关性,且干线长度的影响更为显著,随着干线长度的增加,行程时间趋于一个稳定的单峰分布,波动性减小,可靠性增加。
The travel time distribution and reliability characteristics of urban arterial roads with multiple signalized intersections are one of the important branches of traffic flow theory.Based on the grey relational analysis,this paper establishes a model of travel time influencing factors of multiple signalized intersections under signal coordination control.Firstly,the automatic number plate recognition(ANPR)data is preprocessed to get travel time data;then,the Burr distribution and the Gaussian mixture model are used to fit the travel time distribution and conduct goodness of fit test;finally,the grey correlation method is used to analyze the relationship between the number of signalized intersections,route length,route traffic flow and the travel time characteristics.The results show that the link travel time distribution is bimodal,and the Gaussian mixture model is suitable for the fitting of link travel time;the Burr distribution can better describe the characteristics of right deviation of the arterial roads with multiple signalized intersections.There is a certain correlation between the number of signalized intersections,route length,route traffic flow and travel time characteristics,and the influence of the route length is more significant.With the increase of route length,the travel time tends to a stable unimodal distribution,and the fluctuation decreases and the reliability increases.
作者
王福建
卢一笑
金盛
WANG Fu-jian;LU Yi-xiao;JIN Sheng(College of Civil Engineering and Architecture,Zhejiang University,Hangzhou 310058,China)
出处
《交通运输系统工程与信息》
EI
CSCD
北大核心
2021年第4期132-139,共8页
Journal of Transportation Systems Engineering and Information Technology
基金
国家重点研发计划(2018YFB1601000)
国家自然科学基金(92046011)。
关键词
交通工程
行程时间特性
BURR分布
高斯混合模型
影响因素
traffic engineering
travel time characteristics
Burr distribution
Gaussian mixture model
influence factor
