基于混合低频车辆轨迹的交叉口延误计算  

Calculation of intersection delay based on mixed low-frequency vehicle trajectories

作　　者：张建波 孙建平 李霄 赵祥 徐德中 ZHANG Jianbo;SUN Jianping;LI Xiao;ZHAO Xiang;XU Dezhong(Beijing Key Laboratory of Urban Transport Simulation and Decision Making Support,Beijing Transport Institute,Beijing 100073,China;Beijing International Science and Technology Cooperation Base of Urban Transport,Beijing Transport Institute,Beijing 100073,China;Beijing Zhijia Travel Technology Co.,Ltd.,Beijing 100044,China)

机构地区：[1]北京交通发展研究院城市交通运行仿真与决策支持北京市重点实验室,北京100073 [2]北京交通发展研究院城市交通北京市国际科技合作基地,北京100073 [3]北京智驾出行科技有限公司,北京100044

出　　处：《北京交通大学学报》2024年第4期123-130,共8页JOURNAL OF BEIJING JIAOTONG UNIVERSITY

基　　金：国家自然科学基金(62173167)。

摘　　要：针对道路交通运行监测评价中不同来源的网联车辆轨迹存在采样频率低和采样频率不一致等问题,提出了一种面向混合低频车辆轨迹的交叉口延误融合计算方法.首先,设计分段拟合函数重构网联车辆低频轨迹,再现车辆在交叉口区域的速度变化趋势与加减速临界点.然后,基于实测轨迹量化分析不同采样频率与通过速度条件下交叉口车辆延误的误差特征,并以此为基础构建面向多源轨迹的交叉口延误融合计算与修正方法,利用先验误差修正和加权混合低频轨迹的延误计算结果.最后,通过蒙特卡洛实验验证所提出方法的有效性.研究结果表明:车辆行驶速度和样本量大小均对交叉口低频轨迹车辆延误的误差有显著影响;在相同速度区间和样本量条件下,采样间隔时间越大,重构后低频轨迹车辆延误的平均误差越大;在混合低频轨迹样本量为10的条件下,利用所提出方法计算的交叉口车均延误的相对误差在8%范围内的概率为93.6%,所提出方法能够有效提升利用混合低频轨迹测算交叉口机动车延误的精度.To address the issues of low and inconsistent sampling frequencies of trajectories for intelligent and connected vehicles from different sources in road traffic operation monitoring and evaluation,this paper proposes a method for calculating intersection delay using mixed low-frequency vehicle trajectories.Firstly,a piecewise fitting function is designed to reconstruct low-frequency trajectories of intelligent and connected vehicles,capturing the speed variation trends and critical acceleration/deceleration points within the intersection area.Then,based on measured trajectories,the error characteristics of intersection vehicle delay under different sampling frequencies and passing speeds are quantitatively analyzed.Building on this,a fusion calculation and correction method for intersection delay using multi-source trajectories are developed,utilizing prior error correction and weighted mixed lowfrequency trajectory delay results.Finally,the effectiveness of the proposed method is validated through Monte Carlo experiments.The results show that both vehicle speed and sample size significantly impact the delay error of low-frequency trajectory vehicles at intersections.With the same speed interval and sample size,the larger the sampling interval,the greater the average delay error of reconstructed low-frequency trajectories.With a sample size of 10 mixed low-frequency trajectories,the proposed method achieves a relative error probability of 93.6% within 8% for the average vehicle delay at intersections,demonstrating that the method effectively enhances the accuracy of intersection delay measurements using mixed low-frequency trajectories.

关 键 词：交通工程 交叉口延误 混频数据 低频轨迹 数据融合 

分 类 号：U491[交通运输工程—交通运输规划与管理]