基于DBSCAN算法的住宿业集群空间分布特征及影响因素研究:以乌鲁木齐市为例

Research on spatial distribution characteristics and influencing factors ofaccommodation industry clusters based on DBSCAN algorithm:taking Urumqi City as an example

探讨城市住宿业的空间分布格局,对城市产业格局优化及其有序发展具有重要意义.本文以乌鲁木齐市为例,选取其住宿设施为研究对象,通过高德地图API获取乌鲁木齐市7区1县1948个POI(point of interest)数据,并运用DBSCAN(density-based spatial clustering of applications with noise)算法识别其核心集群,揭示该市住宿业集群空间分布特征及影响因素.研究结果显示:1)乌鲁木齐市住宿业集群分为5个等级,形成了鲜明的“一主一次”的住宿业集群分布格局,且可归纳为基于城区基础设施的综合型和基于城郊自然景观的旅游型2种核心集群发展模式;2)乌鲁木齐市住宿业分布呈“南-北”走向,形成了“一点一带一团一簇”的分布模式;3)商场超市密度、路网密度、公交和地铁站点密度是乌鲁木齐市住宿业集群空间分布的重要影响因素.

Exploring the spatial distribution pattern of urban accommodation industry is of great significance to the optimization of urban industrial pattern and its orderly development.Taking Urumqi City as an example,this paper selects its accommodation facilities as the research object,obtains the data of 1948 POI in seven districts and one county of Urumqi through Gaode Map API,uses DBSCAN algorithm to identify its core clusters,and reveals the spatial distribution characteristics and influencing factors of Urumqi's accommodation industry clusters.The research results show that:1)Urumqi's accommodation industry cluster is divided into five levels,forming a distinctive"one primary and one secondary"distribution pattern of accommodation industry cluster,which can be summarized into two core cluster development models:comprehensive cluster based on urban infrastructure and tourism cluster based on suburban natural landscape;2)The distribution of accommodation industry in Urumqi shows a"south-north"trend,forming a distribution pattern of"one point,one belt,one area,one cluster".3)The density of shopping malls and supermarkets,the density of road network,and the density of bus and subway stations are important factors affecting the spatial distribution of Urumqi's accommodation industry clusters.