基于交通事故间距分布特征的事故黑点鉴定新方法  

New Traffic Accidents Black Spots Identification Method Based on Abnormal Accidents Space

作　　者：崔洪军[1] 申晓静[1] 赵述捷[1] 孙喜欣 

机构地区：[1]河北工业大学土木工程学院,天津300401 [2]河北省高速公路服务管理中心,秦皇岛066000

出　　处：《武汉理工大学学报》2012年第2期54-58,共5页Journal of Wuhan University of Technology

基　　金：国家自然科学基金(50808064)

摘　　要：常用的事故多发点鉴定方法都需在分析事故数据前将路段进行划分,这与事故的空间分布不一致,易导致事故多发点的割裂分析。在对交通事故数据统计分析基础上,验证了单位里程事故数服从泊松分布,并推导出事故间距服从负指数分布,进而可计算出异常事故间距的界限值,认为构成异常事故间距的事故与路况可能存在一定联系,并根据小概率事件原理,可确定出异常事故间距连续出现个数的下限值,大于该下限值的连续事故异常间距构成路段即为事故多发地点。Commonly-used method on black-spot appraisal needs to division the section before the analysis of the data,which is not agree with the spatial distribution of the accident,leading to the black-spot separate analysis.This paper,based on the analysis of the traffic accidents data,verify that the number of accidents in unit mileage obey Poisson distribution and infer that spaces between traffic accicents obey negative exponent distribution,then the abnormal space of traffic accicents can be found out.Thinking the abnormal space accident may have some relation with road conditions,and according to the principle of small probability events,determine the lower limit of abnormal spacing number,if the number of abnormal spacing in one section more than the limit,then take this section as traffic accidents black spots.

关 键 词：交通事故 事故多发点 事故间距 泊松分布 

分 类 号：U491[交通运输工程—交通运输规划与管理]