摘 要
购物中心作为城市居民购物和休闲的重要载体，其空间分布不但影响着自身的经营
发展也对城市空间结构产生巨大影响。
首先，本文基于 2008年-2021年西安市购物中心统计数据，认为西安市购物中心发
展经历了初步发展阶段、快速发展阶段和稳健发展阶段；利用 2013年、2018年、2021
年三个时间截面购物中心 POI数据与莫兰指数、核密度分析法，归纳购物中心空间演化
特征：整体空间分布呈现空间集聚度不断增强、集聚区向多极演化、集聚区显现两级结
构、沿交通主干线分布特征。局域空间分布显示“H-H”活力区域集中分布于城市中心和
城市次中心、“L-H”空心区分布于“H-H”活力区的周边、“H-L”极化区分布于城市外围、
“L-L”滞后区分布在城市外围的特征。
其次，利用 43家购物中心营业额数据和 K均值聚类法对购物中心进行等级划分，
分为市级购物中心、区域级购物中心和社区级购物中心三级。
最后，利用购物中心 POI数据，通过自变量检验、空间效应检验以及模型拟合检验
选择空间误差模型探究购物中心空间布局的影响机理，发现人口数量、城市规划、交通
区位、居住社区、景点、房价水平在不同方面对购物中心空间演化影响作用差异较大。
①从购物中心纵向发展角度分析，起步发展阶段，与城市中心距离、零售繁荣度对购物
中心分布的影响最为显著；快速发展阶段，城市商业规划与地铁网络对购物中心空间分
布影响较为突出；稳健发展阶段，地铁网络对购物中心空间分布影响最为显著。②从购
物中心空间分布角度分析，零售繁荣度对城市内圈层购物中心分布影响较为突出，而城
市商业规划和轨道交通是影响城市外圈层购物中心的分布的核心要素。③从购物中心等
级结构角度分析，商业规划和地铁网络是影响高等级购物中心核心因素；景点显著影响
区域级购物中心分布；居住社区是影响社区级购物中心分布重要因素。
关键词：购物中心，空间分布，等级结构，演化机理
i
Abstract
As an important carrier of shopping and leisure for urban residents, the spatial distribution
of shopping centers not only affects their own business development but also has a great impact
on the spatial structure of the city.
Based on the statistical data of shopping centers in Xi'an from 2008 to 2021, this paper
argues that the development of shopping centers in Xi'an has gone through a preliminary
development stage, a rapid development stage and a steady development stage; using the POI
data of shopping centers in three time sections of 2013, 2018 and 2021 with the Moran index
and the nuclear density analysis method, we summarize the spatial evolution characteristics of
shopping centers: the overall spatial distribution shows The overall spatial distribution shows
increasing spatial agglomeration, multi-polar evolution of agglomerations, two-level structure
of agglomerations, and distribution along the main traffic routes. The local spatial distribution
shows that the "H-H" dynamic areas are concentrated in urban centers and sub-centers, while
the "L-H" hollow areas are distributed around the "H-H" dynamic areas. The "H-L" polarization
zone is distributed in the periphery of the city, and the "L-L" lagging zone is distributed in the
periphery of the city.
Secondly, using the turnover data of 43 shopping centers and the K-means clustering
method, the shopping centers were classified into three levels: city-level shopping centers,
regional-level shopping centers and community-level shopping centers.
Finally, using the shopping center POI data, the spatial error model was selected by the
independent variable test, spatial effecttest and model fit test to explore the influence
mechanism of the spatial layout of shopping centers, and it was found that the population size,
urban planning, traffic location, residential community, attractions, and housing price level
differed greatly in their effects on the spatial evolution of shopping centers in different aspects.
①Analyzing from the perspective of vertical development of shopping centers, the distance to
the city center and retail prosperity have the most significant influence on the distribution of
shopping centers in the initial development stage; in the rapid development stage, urban
commercial planning and metronetwork have more prominent influence on thespatial
distribution of shopping centers; in the steady development stage, metro network has the most
ii