考虑建成环境的酒驾事故时空分布特征
Consider the spatial and temporal distribution characteristics of drunk driving accidents in the built environment
于晓桦 1田雨晗 1杨大志1
作者信息
- 1. 山东建筑大学 交通工程学院,济南 250101
- 折叠
摘要
为深入了解我国酒驾事故的时空分布特征及其与周围建成环境的空间距离阈值,以济南市 2022 年发生的1 034 起酒驾事故与13 029起非酒驾事故案例为研究对象,结合季节性强度指数与地理集中指数等衡量指标,分析酒驾事故的时空分布特征;利用近邻分析与基于最大类簇直径改进的K-means聚类算法,探究出周围建成环境与酒驾事故点的空间距离阈值.结果表明:酒驾事故的时间分布季节强度指数为 4.128,大于非酒驾事故的 3.839,时间分布差异性更大,且高发时段(20:00-02:00)与餐饮服务的高峰营业时间一致.酒驾事故的地理集中指数 G为32.713%,大于均匀分布情况下的地理集中指数(G),空间集聚程度更高,且与周围建成环境分布具有密切联系.餐饮服务、购物服务、住宿服务与商务住宅的聚类中心点到酒驾事故点的距离阈值分别为 798.415 m、694.378 m、658.234 m与 739.622 m.
Abstract
In order to deeply understand the spatiotemporal distribution characteristics of drunk driving accidents in China and the spatial distance threshold from the surrounding environment,taking 1034 drunk driving accidents and 13 029 non-drunk driving traffic accidents in Jinan in 2022,this paper combines measures such as seasonal intensity index and geographic concentration index,analyze the spatial and temporal distribution characteristics of drunk driving accidents and use a close-neighbor analysis and an improved K-means clustering algorithm based on the maximum-class cluster diameter.It explores the spatial distance threshold between the surrounding built environment and the drunk driving accident point.The results show that the seasonal intensity index of the time distribution of drunk driving accidents is 4.128,which is greater than the 3.839 of non-drunk driving accidents,and the difference of the time distribution is greater,and the high incidence period(20:00-02:00)is consistent with the peak business time of catering service.The geographic concentration index of drunk driving accidents G=32.713%,which is greater than the geographic concentration index(G)under uniform distribution,has a higher degree of spatial agglomeration and is closely related to the distribution of the surrounding environment.The distance threshold between catering service,shopping service,accommodation service and business residence is 798.415 m,694.378 m,658.234 m and 739.622 m,respectively.
关键词
交通安全/酒驾事故/近邻分析/K-means聚类算法/距离阈值Key words
traffic safety/drunk driving accident/nearest neighbor analysis/K-means clustering algorithm/distance threshold引用本文复制引用
出版年
2024
NETL
NSTL
万方数据