核密度法的南京苏果超市分布热点探测  

Hotspot detection of Suguo stores in Nanjing city supported by kernel density estimation

作　　者：吴嘉逸 席唱白 苑振宇[1] 芮一康[1] 王结臣[1,2] 

机构地区：[1]南京大学地理信息科学系,南京210023 [2]江苏省地理信息技术重点实验室,南京210023

出　　处：《测绘科学》2017年第11期68-73,共6页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41571377);国家自然科学基金青年项目(41401450)

摘　　要：针对超市经营者进行选址优化和空间配置问题,该文以南京市主城区内苏果超市为例,利用平面核密度方法,对城市内零售超市的空间结构与分布模式进行分析,系统地研究不同类型超市的空间分布特征并进行热点区域的探测。研究表明,热点区域的探测结果符合商业网点分布的一般规律,南京主城区内不同类型苏果超市的空间分布表现出较大的空间区域分异:购物广场的热点分布于中心城区外围的次一级中心,社区店和标准超市的热点分布于居住区,便利店的热点大多分布于人口密集区域。The analysis of spatial structure and distribution pattern of retail stores in city contributes to optimization of location selection and space allocation.This research takes Suguo stores in the main area of Nanjing as an example,using the method of plane kernel density estimation to study the spatial distribution characteristics of different types of stores systematically and detect hotspots.This study shows that the detection results of the hotspots are in accordance with the general rules of the distribution of commercial network.The spatial distribution of different types of stores in the main area of Nanjing shows a large spatial regional differentiation：the hotspots of hypermarkets locate at sub-level centers around the central city.The hotspots of community stores and standard stores locate at residential areas.The hotspots of convenience stores locate at areas with dense population mostly.Results show that the method of plane kernel density estimation is able to effectively distinguish the difference in hotspots distribution which is caused by different types.

关 键 词：核密度估计 k阶最邻近法 苏果超市 热点探测 

分 类 号：P208[天文地球—地图制图学与地理信息工程]