文章编号： 1672-5328（2018）03-0021-07

朱新宇，丛浩哲，支野，索子剑
(公安部道路交通安全研究中心，北京100062)

摘要： 城市道路交通事故黑点及事故多发路段是城市交通管理中的顽疾所在，但如何在现有数据基 础上精准识别和定位事故多发路段尚无统一定论。以深圳市为例，基于2014—2016 年交通事故数 据，利用空间模糊度搜索调用数字地图API 来确定事故点位，并基于累计频率分析法对道路进行事 故黑点分析，最后利用GIS 线性参考技术对事故点位进行精确定位。结合GIS 平台进行系统封装和 二次开发，构建一套基于GIS 空间聚类的城市道路交通事故多发路段鉴别分析系统。应用实例显 示，系统可实现城市道路交通事故多发路段的精准定位及分析等多项功能。

关键词： 交通安全；事故多发路段；事故黑点；空间聚类；大数据；GIS

中图分类号： U491.3

文献标识码：A

Accident-Prone Location Analysis Based on GIS Spatial Clustering

Zhu Xinyu, Cong Haozhe, Zhi Ye, Suo Zijian
(Road Traffic Safety Research Center of the Ministry of Public Security, Beijing 100062, China)

Abstract： Urban road traffic accident black spots and accident-prone locations are chronically problematic for urban traffic management. However, there is no consensus on how to accurately identify and locate the accident- prone roadway locations with the existing data. Based on the traffic accident data in Shenzhen from 2014 to 2016, this paper identifies the accident locations using space fuzzy search for the digital map API. Through analyzing the black spots using cumulative frequency method, the paper introduces the method that can identify accident accurate location based on GIS linear referencing technology. An identification and analysis system for accident-prone location based on GIS spatial clustering is developed by original system package and secondary program development on GIS platform. Application examples show that the system can obtain precise positioning and analysis of accident-prone locations among other functions.

Keywords： traffic safety; accident-prone locations; accident black spots; spatial clustering; big data; GIS