基于聚类分析的道路交通事故多发点段评价方法  

A Cluster Analysis Based on Evaluation Method for Road Traffic Accidents Prone

作　　者：常安德 张朝旭[2] 陈松 CHANG Ande;ZHANG Zhaoxu;CHEN Song(Key Laboratory of Impression Evidence Examination and Identification Technology,Shenyang 110035,China;Criminal Investigation Police University of China,Shenyang 110035,China)

机构地区：[1]公安部痕迹检验鉴定技术重点实验室,辽宁沈阳110035 [2]中国刑事警察学院,辽宁沈阳110035

出　　处：《中国人民公安大学学报（自然科学版）》2020年第2期47-52,共6页Journal of People’s Public Security University of China(Science and Technology)

基　　金：道路交通安全公安部重点实验室开放课题(2019ZDSYSKFKT07);痕迹检验鉴定技术公安部重点实验室资助课题(20151205);中国刑事警察学院研究生创新能力提升项目(2019YCYB27)。

摘　　要：鉴于道路交通事故成因的多样性和复杂性,在定义道路交通事故多发点段和充分考虑交通事故空间定位信息的基础上,基于聚类分析技术研究了一种道路交通事故多发点段的识别方法,同时设计一种事故多发点段空间范围的动态非定长划分方法,进而利用当量总事故率法将交通事故多发点段划分为3个风险等级,并给出了具体的划分标准,从而为交通事故预防工作提供决策支持。Given the diversity and complexity of road traffic accident formation,the definition of road traffic accident black point and give full consideration to the traffic accident,on the basis of spatial location information,based on the clustering analysis to study a method of road traffic accident black point recognition,design a kind of accident black point at the same time period of space dynamic non fixed-length division method,and then using the method of equivalent total accidents will traffic accident black point section is divided into three level of risk,and gives the specific division standard,so as to provide decision support for traffic accident prevention work.

关 键 词：事故多发 聚类分析 动态非定长 风险等级 

分 类 号：D631.5[政治法律—政治学]