考虑相关性的雨天城市道路交通拥堵概率预测  

Probability Prediction of Urban Road Congestion on Rainy Days Considering Correlation

作　　者：吴中[1] 张津玮[1] 杨海飞 吴琼[1] WU Zhong;ZHANG Jinwei;YANG Haifei;WU Qiong(College of Civil and Transportation Engineering, Hohai University, Nanjing 210098,China)

机构地区：[1]河海大学土木与交通学院,江苏南京210098

出　　处：《贵州大学学报（自然科学版）》2020年第3期93-97,104,共6页Journal of Guizhou University:Natural Sciences

基　　金：国家自然科学基金项目资助(71801080);江苏省博士后科研项目资助(2018K043B)。

摘　　要：雨天城市道路交通拥堵频发,交通运行在道路网络各路段间的交互影响较晴天明显增强。本文针对传统出行时间预算模型应用的局限性,考虑雨天各路段间行车延误的相互影响,改进了走行时间方差的计算方法,提出一种雨天城市道路交通拥堵概率预测方法。选取南京市某行车路径进行验证,通过4种工况、3种模型走行时间的比较,发现考虑相关性的分段组合模型更接近整体路径分布情况,且拥堵概率预测准确率大于90%。研究结果表明:本文提出的预测模型能有效预测城市道路雨天拥堵情况,可为雨天道路适应性评价和交通管控提供依据。Urban road traffic congestion occurs frequently in rainy days,and the interactive influence of traffic running in each section of the road network is significantly enhanced compared with sunny days.Aiming at the limitation of traditional travel time budget model,this paper improves the calculation method of travel time variance,and proposes a probability prediction model on rainy days congestion considering the correlation between sections.The results of a case study of a path in Nanjing show that compared with the travel time under four conditions and three models,the piecewise combination model considering correlation is closer to the travel time distribution of the global path model,and the prediction precision of congestion probability is more than 90%.The results show that this prediction model can effectively predict urban road congestion on rainy days,which can provide a theoretical basis for road adaptability evaluation and traffic control on rainy days.

关 键 词：交通运输系统工程 拥堵概率预测 雨天 走行时间 路段相关性 

分 类 号：U491[交通运输工程—交通运输规划与管理]