基于数据采样方法的城市道路自由流行程时间估计  

Free-flow travel time estimation in urban roads based on data sampling method

作　　者：俞怡 曾佳棋 王殿海[1] YU Yi;ZENG Jia-qi;WANG Dian-hai(College of Civil Engineering and Architecture,Zhejiang University,Hangzhou 310058,China)

机构地区：[1]浙江大学建筑工程学院,浙江杭州310058

出　　处：《浙江大学学报（工学版）》2022年第6期1241-1248,共8页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(52131202,61773338,52072340);山东省重大科技创新工程项目(2019TSLH0203)。

摘　　要：基于交通波模型,提出利用车牌识别数据估计城市道路自由流行程时间.无需额外架设检测器或现场测算,所提方法具备准确性、科学性、实用性的特点.基于车辆均匀到达的假设,将行程时间分为自由流行程时间和延误,建立信号影响下的路段行程时间分布函数.针对现实环境中车流非均匀到达的特点,提出数据重采样方法生成符合均匀流假设的行程时间数据;拟合行程时间分布函数以获得路段自由流行程时间.在杭州市多个路段的数据验证结果表明,重采样后的行程时间数据较好地拟合了行程时间分布模型,估得的自由流行程时间准确且具备理论支撑.Based on traffic wave theory,free-flow travel time in urban areas was estimated using plate recognition data.The proposed method was accurate,scientific and practical with no need for additional detectors or on-site calculations.With the hypophysis of uniform arrival,travel time was divided into free-flow travel time and delay,and a travel time distribution function was constructed in the traffic signal environment.To solve the non-uniform arrival problem in urban traffic,a data resampling method was proposed to generate travel time data that obey the uniform flow assumption.By fitting the travel time distribution function,the free-flow travel time was estimated.The approach was verified in Hangzhou,China.Results showed that the travel time data after resampling fits the travel time distribution model well.The estimated free-flow travel time is accurate and has fully theoretical support.

关 键 词：自由流行程时间 行程时间分布 数据重采样 城市道路 

分 类 号：U491[交通运输工程—交通运输规划与管理]