摘要
事故数据记录不完善制约了基于事故数据的交叉口安全分析方法的运用.冲突数据样本量大、观测周期短,可以作为事故数据的补充,评价交叉口安全状况及分析事故致因.选取上海市区交叉口进行冲突视频观测,提取冲突类型及冲突严重程度等数据.建立了事故率与机-机、机-非、机-人严重冲突率多元线性回归方程,该方程可用于交叉口事故多发点判别.以武宁路-兰溪路交叉口为例进行致因分析,分析冲突类型及分布进而提出针对的改善措施.
In China,crash data analysis for intersection safety cannot be applied,largely because of the limitation of crash data record.Abundant traffic conflict data can be obtained in short period of time and therefore,traffic conflict data can severe as supplement of crash data to evaluate intersection safety condition and analyze crash causes.Traffic conflicts of intersections in urban areas of Shanghai were observed using cameras and conflict pattern and severity were extracted.Multiple linear regression equation was build to relate crash rate with vehicle-vehicle,vehicle-bicycle and vehicle-pedestrian serious conflict rate.It can be used to identify intersection hotspot.Taking Wuning-Lanxi intersection as an example of crash cause analysis,conflict pattern and location were analyzed and count-measures for safety improvement were put forward.
出处
《武汉理工大学学报（交通科学与工程版）》
2012年第2期370-373,共4页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目(批准号:51008230)
同济大学道路与交通工程教育部重点实验室自主课题项目资助
关键词
交通冲突
交叉口安全
事故多发点鉴别
多元线性回归
致因分析
traffic conflict
intersection safety
hotspot identification
multiple linear regression
crash cause analysis
