文献类型：期刊文献

中文题名：混行交通流中两轮电动车换道行为分析

英文题名：Analysis of Lane-changing Behaviors of Electric Two-wheelers in Mixed Traffic Flow

作者：葛鹏 陈梅

第一作者：葛鹏

机构：[1]贵州理工学院交通工程学院,贵阳550025;[2]贵州财经大学发展规划处,贵阳550025

第一机构：贵州理工学院交通工程学院

年份：2025

卷号：22

期号：2

起止页码：63-68

中文期刊名：现代交通技术

外文期刊名：Modern Transportation Technology

语种：中文

中文关键词：交通管理;混行交通流;关联分析;二元逻辑回归模型

外文关键词：traffic management;mixed traffic flow;correlation analysis;binary logistic regression model

摘要：为探究两轮电动车换道行为与动静态交通环境之间的关系,基于贵阳市3条城市主干道,通过无人机航拍与人工提取相结合的方式采集电动车换道数据,并定义了13种动静态交通环境。采用关联规则挖掘算法,分析两轮电动车换道行为与交通环境的关联性,在支持度>0.1、置信度>0.5的条件下,得到15条关联规则。规则表明,在交通环境因素中,前方有慢车或停车以及公交车出站是两轮电动车换道的强力影响因素。此外,构建了两轮电动车换道行为的二元逻辑回归模型,结果表明:前方有慢车或停车、车辆违停、公交车出站、车辆驶入道路等因素对换道影响较大。研究结果可辅助城市道路风险路段的识别与预测工作,同时,对交通设施设置与交通管理工作具有借鉴意义。
To investigate the relationship between lane-changing behavior of electric two-wheels and dynamic/static traffic environments,lane-changing data was collected using drone aerial photography followed by manual extraction on three main urban roads in Guiyang City.Thirteen types of dynamic and static traffic environments were defined.The association rule mining algorithm was employed to analyze the correlation between electric two-wheeler lane-changing behavior and traf-fic environments.Under the conditions of support>0.1 and confidence>0.5,15 association rules were obtained.The rules indicate that among the traffic environmental factors,slow or parked vehicles ahead and buses exiting stops are strong influencing factors for electric two-wheeler lane-changing.Additionally,a binary logistic regression model was con-structed for electric two-wheeler lane-changing behavior.The results show that factors such as slow or parked vehicles ahead,illegally parked vehicles,buses exiting stops and vehicles entering the road have significant impacts on lane-changing.The findings can assist in identifying and predicting high-risk road sections in urban areas,while also provi-ding valuable insights for traffic facility planning and management.