摘 要

变道行为是最基本的车辆行驶行为之一，但不当的变道行为会造成危险，引发交通事故。信号交叉口是交通网络车辆改变行驶方向的交通节点，此处交通流量大，交通环境复杂，经常发生车辆变道的情况，部分驾驶员为了抢先驶入目标车道或排队较短的车道，往往在进口道采取危险的变道行为。一方面会扰乱原来有序的交通流，降低交叉口的通行效率；另一方面会侵占周围车辆的行驶空间，增加交通冲突的数量，增大事故发生的概率。本文基于无人机航拍的交叉口视频数据，通过分析车辆时变轨迹对变道行为的危险性进行了研究。

本研究选取武汉市典型信号交叉口进行人工观测及无人机视频观测，运用视频处理软件Tracker分析视频，提取视频中危险变道车辆以及周围车辆的坐标、速度等，分析处理变道车辆时变轨迹上各点的交通流参数，由此给出了信号交叉口危险变道行为的定义，提出了危险变道行为分类方法；对危险变道行为进行了阶段划分，并基于碰撞时间（）提出变道碰撞时间（）的概念，建立轨迹上各点对应方向（与行车道平行）及方向（与行车道垂直）上的变道碰撞时间模型；基于变道行为的三阶段及每阶段对应的变道碰撞时间平均值，建立基于时变轨迹的变道行为危险性识别模型，运用熵权法确定参数权重，并对危险度进行分级；基于视频数据的分析及危险性识别模型的应用，以友谊大道-工业路交叉口为研究对象，分析信号交叉口变道行为的危险性，包括变道行为不同阶段特性、不同类型变道行为特性，并在分析结果的基础上提出信号交叉口进口道危险变道行为改善措施。

关键词：信号交叉口；变道行为；时变轨迹；危险性；变道碰撞时间

Abstract

The lane change behavior is the driving behavior that often occurs at signalized intersections. However, improper lane change will cause danger and traffic accidents. The signalized intersection is the traffic node where the transportation network vehicle changes the driving direction. There are large traffic flow, complex traffic environment, traffic jams and frequent lane change behaviors at the signalized intersection. Some drivers often make dangerous lane changes in order to drive into the target lane or the lane with short queue faster. Such dangerous lane changes will disrupt the orderly traffic flow and reduce the traffic efficiency at intersections. On the other hand, they will invade the driving space of surrounding vehicles and increase the probability of conflict and accidents. Based on the intersection video data from UAV aerial photography, we study the risk of lane change behaviors based on the time-varying trajectory.

In this paper, we selected Wuhan's typical signalized intersection for manual observations and UAV video observations, using a video processing software (Tracker) to analyze the video. From the video, we can extract and analyze the coordinate and speed of lane-changing vehicles and surrounding vehicles. As a result, we defined and classified the dangerous lane change behavior. Additionally, we divided the dangerous lane change behavior and proposed the prospect of lane change collision time () based on collision time ().We established the lane change collision time model at the directions parallel to the roadway () and perpendicular to the traffic lane (). In view of the three phases’ average lane change collision time, we established a risk model based on the time-varying trajectory. The weights were determined by entropy weight method and the risk degree is graded; Through the analysis of video data and the use of risk model, we analyzed the risk of lane change behaviors at the intersection of Youyi Avenue-Industry Road, including the characteristics of different types of lane change behaviors and different stages of lane change behaviors. Based on the analysis results, the measures to improve the behavior of dangerous lane changes at the entrances to signalized intersections were proposed.

Keywords: Signalized intersection; Lane change behavior; Time-varying trajectory; Risk; Lane change collision time

目 录

目 录 1

第1章 绪论 1

1.1研究背景 1

1.2研究目的及意义 1

1.2.1研究目的 1

1.2.2研究意义 1

1.3国内外研究现状 2

1.3.1驾驶员变道行为研究 2

1.3.2驾驶员变道模型研究 2

1.3.3交通冲突的危险性研究 2

1.3.4研究现状评述 3

1.4研究内容与技术方案 3

1.4.1研究内容 3

1.4.2研究方法 4

1.4.3研究技术路线 4

1.5论文结构 6

第2章 车辆危险变道行为数据调查 7

2.1危险变道行为定义 7

2.2危险变道行为分类 7

2.3调查目的与内容 8

2.3.1调查目的 8

2.3.2调查内容 8

2.4调查方案 8

2.4.1调查方法 8

2.4.2调查地点 10

2.4.3调查流程 12

2.5调查数据提取 12

2.5.1样本提取原则 12

2.5.2数据提取方法 12

2.5.3提取主要参量 13

2.6调查数据统计分析 14

2.6.1危险变道类型规律 14

2.6.2危险变道车型分布规律 15

2.6.3危险变道车辆速度时变特性 15

2.6.4危险变道车辆位置时变特性 18

2.7本章小结 20

第3章 信号交叉口变道行为危险性识别模型 21

3.1车辆变道阶段划分 21

3.2模型研究思路 21

3.3变道行为时变轨迹冲突点的危险性识别模型 22

3.3.1 碰撞时间（）基本原理 22

3.3.2基于变道碰撞时间（）的危险性识别模型 22

3.4变道行为危险性识别模型 24

3.4.1熵权法介绍 25

3.4.2熵权法基本步骤 25

3.4.3熵权法赋权 26

3.5危险度等级划分 27

3.6本章小结 27

第4章 信号交叉口变道行为危险性分析 29

以上是毕业论文大纲或资料介绍，该课题完整毕业论文、开题报告、任务书、程序设计、图纸设计等资料请添加微信获取，微信号：bysjorg。

相关图片展示：