基于航拍数据的互通隧道小净距路段运行速度模型  

Operating speed model of tunnel exit and interchange exit in short distance sections based on aerial photography data

作　　者：张驰[1] 高艳阳 杨榕玮 刘昌赫 谢子龙 靳引利[3] ZHANG Chi;GAO Yan-yang;YANG Rong-wei;LIU Chang-he;XIE Zi-long;JIN Yin-li(School of Highway,Chang'an University,Xi'an 710064,Shaanxi,China;Shaanxi Provincial Transport Planning Design and Research Institute,Xi'an 710065,Shaanxi,China;School of Electronics and Control Engineering,Chang'an University,Xi'an 710064,Shaanxi,China)

机构地区：[1]长安大学公路学院,陕西西安710064 [2]陕西省交通规划设计研究院,陕西西安710065 [3]长安大学电子与控制工程学院,陕西西安710064

出　　处：《长安大学学报（自然科学版）》2024年第2期136-150,共15页Journal of Chang’an University(Natural Science Edition)

基　　金：国家重点研发计划项目(2020YFC1512005,2019YFB1600700);四川省科技计划项目(2022YFG0048);藏高科技项目(2021-11);陕西省自然科学基础研究计划项目(2023-JC-YB-391)。

摘　　要：为探究车辆在隧道出口-互通出口小净距路段运行速度变化规律,改善当前山区高速公路小净距路段行车安全性不足现状,开展小净距路段车辆运行速度模型研究。首先,基于无人机航拍获取西汉(西安—汉中)高速等4处隧道出口-互通出口小净距路段车辆实测数据,利用YOLO-v5算法编写航拍视频数据提取程序,输出车辆在目标路段的实时速度数据和轨迹数据,同时结合驾驶仿真模拟系统对试验进行对照组数据补充。其次,统计车辆在该区域运行情况及断面速度分布并进行宏观分析,确定构建运行速度模型的自变量。最后,应用非线性曲线拟合软件1stOpt对分析数据进行回归,分别建立小型车、大型车在出口小净距路段主线行驶、驶出主线的运行速度预测模型,并依托西南山区某小净距工程实际路段开展预测模型的有效性验证。研究结果表明:出口小净距路段车辆运行速度与主线半径、车辆距隧道洞口距离和净距长度存在显著正相关关系;建立的运行速度预测模型通过了拟合优度检验、统计学残差检验和相对平均误差检验,模型预测值与实测值的相对误差平均值均小于10%,回归模型满足精度要求。本研究能应用于实际工程中的隧道出口-互通出口小净距路段,能准确表征车辆在此区域的速度变化规律,在预测车辆运行速度和表征速度变化规律方面具有一定的参考价值。To explore the pattern of vehicle speed variation in short distance sections between tunnel exits and interchange exits,and to alleviate the current insufficiency in driving safety on short-distance sections of mountainous highways.Vehicle speed models in such sections was studied,Firstly,based on aerial drone footage,real-time vehicle speed and trajectory data were extracted from sections with short clearance between tunnel exits and interchange exits on the Xi-Han(Xi'an to Hanzhong)Expressway and three other locations,and a program written with the YOLO-v5 algorithm was used.This was complemented with control group data from a driving simulation system.Secondly,the vehicle operation and sectional speed distribution in this area for macroscopic analysis were analyzed statistically,the independent variables for constructing the speed model was determined.Finally,non-linear curve fitting software 1stOpt was applied to regress the analyzed data,establishing predictive models for the operational speeds of small and large vehicles driving on or exiting the mainline in short-distance exit sections.These models were validated in a real-world short-distance section project in a Southwestern Mountainous Area.The results show that there is a significant positive correlation between vehicle speed in shortdistance exit sections and mainline radius,the vehicle's distance from the tunnel entrance,and the clearance length.The established speed prediction models passed tests for goodness of fit,statistical residual examination,and relative average error,with an average relative error between predicted and actual values of less than 1o%,satisfying accuracy requirements.This research can be applied to actual engineering projects in short-distance sections between tunnel exits and interchange exits,accurately characterizing vehicle speed variations in this area and providing valuable reference for predicting vehicle operational speeds and representing speed variation patterns.

关 键 词：道路工程 运行速度模型 YOLO-v5算法 小净距路段 航拍数据 

分 类 号：U411[交通运输工程—道路与铁道工程]