基于DBSCAN算法的北京市顺丰快递服务设施集群识别与空间特征分析
张 亚(1995— ),女,甘肃会宁人,硕士生,主要从事空间数据挖掘及分析研究。E-mail:zhangya1549@163.com |
收稿日期: 2019-07-17
要求修回日期: 2019-11-25
网络出版日期: 2020-10-25
基金资助
中国测绘科学研究院基本科研业务费项目(AR1904)
国家重点研发计划项目(2016YFC0803106)
国家重点研发计划项目(2017YBF0503601)
兰州交通大学基金平台支持(201806)
版权
Cluster Identification and Spatial Characteristics Analysis of Shunfeng Express Service Facilities based on the DBSCAN Algorithm in Beijing
Received date: 2019-07-17
Request revised date: 2019-11-25
Online published: 2020-10-25
Supported by
The Basal Research Fund of Chinese Academy of Surveying and Mapping(AR1904)
National Key Research and Development Program of China(2016YFC0803106)
National Key Research and Development Program of China(2017YBF0503601)
Funded by Lanzhou Jiaotong University Excellent Platform(201806)
Copyright
电子商务跨越式发展为快递物流行业注入了新鲜血液促使国民经济达到新的增站点,服务网点作为连接快递企业和用户之间的桥梁纽带,逐渐成为城市地理和物流地理的重要研究对象。本文以北京市顺丰快递服务网点为研究对象,首次将DBSCAN聚类算法和无人值守的智能快递柜引入城市物流快递行业研究中,综合使用核密度分析、Ripley's K函数等空间点模式分析方法,定量对比分析有人值守的合作网点和无人值守的智能快递柜两类顺丰快递服务网点的空间布局、集聚特征及影响因素。结果表明:① 基于密度的DBSCAN聚类算法能够快速有效地识别出任意形状的快递服务网点集群,算法识别出24个智能快递柜集群,14个合作网点集群;② 顺丰快递服务网点高密度区主要集中在人口密度大、经济繁华、交通便利的居住区和包含热门商圈的职住区附近,如双井、金融街、三里屯、学院路等;③ 合作网点和智能快递柜两类服务网点均呈集聚性分布,但集聚规模各不相同,具体表现为快递柜集聚规模明显大于合作网点,而集聚强度却小于合作网点;④ 智能快递柜集群密度大,服务半径小,更倾向服务于步行可达范围内的居住小区;合作网点集群密度小,服务半径大,服务对象随服务半径扩展至周边各大职区,对交通可达性的要求更高。⑤ 顺丰快递服务网点布局是地区经济水平、人口规模、交通状况、土地利用类型及城市功能区定位等多种因素综合作用的结果。
张亚 , 刘纪平 , 周亮 , 王勇 , 李鹏飞 . 基于DBSCAN算法的北京市顺丰快递服务设施集群识别与空间特征分析[J]. 地球信息科学学报, 2020 , 22(8) : 1630 -1641 . DOI: 10.12082/dqxxkx.2020.190380
The leap-forward development of e-commerce has injected fresh blood into the express delivery industry to promote the national economy. As a bridge between express delivery companies and users, service outlets have gradually become an important research object of urban geography and logistics geography.In this paper, by taking Shunfeng Express service outlets of Beijing as the research object,we introduced for the first time the DBSCAN clustering algorithm andunattended intelligent express cabinets into the urban logistics express industry. Spatial analysis methods such as the nuclear density analysis and Ripley's K function wereused to quantitatively compare and analyze the spatial pattern, agglomeration features, and influencing factors of the two types of Shunfeng Express service outlets, namely, manned cooperative outlets and unattended intelligent express cabinets.Results show that: (1) The density-based DBSCAN clustering algorithm can quickly and efficiently identify clusters of express service outlets of any arbitrary shape. The algorithm identified 24 intelligent express cabinet clusters and 14 cooperative network clusters. (2) The high-density area of Shunfeng Express service outlets was mainly concentrated in residential areas with large population density, economic prosperity, convenient transportation, and residential areas in popular business districts, such as Shuangjing, Financial Street, Sanlitun, and Xueyuan Road. (3) The spatial distribution of the two types of service outlets, namely, cooperative outlets and smart express cabinets, was in an agglomeration mode, but the scale of agglomeration was different. The scale of express cabinet agglomeration was significantly larger than that of cooperative outlets, while the intensity of agglomeration was smaller than that of cooperative outlets. (4) The intelligent express cabinet had a large cluster density and a smaller service radius, and was more inclined to serve residential areas within the walking distance; the cooperative network had a smaller cluster density and a larger service radius, and the service object extended to the surrounding major areas with the service radius, at the same time, the demand for traffic accessibility increased. (5) The layout of Shunfeng Express service outlets were the result of a combination of the regional economic level, population size, traffic conditions, land use types, and urban functional area positioning.
表1 2015年北京市主城区快递自提点数量分布Tab. 1 Numbers of express delivery points in the main urban areas of Beijing in 2015 |
面积/km2 | 人口/万人 | 人口密度/(人/km2) | POI数量/个 | |||
---|---|---|---|---|---|---|
合作点 | 智能柜 | 总数 | ||||
朝阳区 | 455 | 395.5 | 8692 | 219 | 897 | 1116 |
海淀区 | 431 | 369.4 | 8571 | 136 | 640 | 776 |
西城区 | 51 | 129.8 | 25 451 | 69 | 142 | 211 |
东城区 | 42 | 90.5 | 21 548 | 37 | 103 | 140 |
丰台区 | 306 | 232.4 | 7595 | 71 | 526 | 597 |
石景山区 | 84 | 65.2 | 7762 | 19 | 112 | 131 |
表2 北京市顺丰快递服务网点依托类型Tab. 2 Types of Shunfeng express service outlets in Beijing |
网点类型 | 智能柜 | 合作点 | ||||||
---|---|---|---|---|---|---|---|---|
依托方式 | 丰巢智能柜 | 便利店、超市 | 物流及快递公司 | 校园 | 个人销售 | 物业及社区服务站 | ||
数量/个 | 2420 | 235 | 108 | 40 | 34 | 134 | ||
占比/% | 81.45 | 7.91 | 3.64 | 1.35 | 1.14 | 4.51 |
图3 北京市顺丰快递服务网点集群分析结果Fig. 3 Catering clusters and the Kernel Density analysis results of Shunfeng Smart-box outlets |
表3 北京市顺丰快递服务网点集群特征Tab. 3 Cluster characteristics of Shunfeng express service outlets |
类别 | 等级/集群 | 集群数量/个 | 紧凑度 | 密度/(个/km2) | 地点 |
---|---|---|---|---|---|
智能快递柜 | 1/3 | 513 | 32.39 | 4.87 | 建外、六里屯、朝外、双井等30个街道 |
2/5 | 188 | 19.46 | 5.32 | 白纸坊、广安门外、太平桥、南苑等10个街道 | |
3/8 | 159 | 17.41 | 5.35 | 田村路、八里庄、八里庄、紫竹院等8个街道 | |
4/16 | 137 | 15.25 | 5.84 | 安贞、小关、大屯、亚运村4个街道 | |
5/7 | 94 | 10.45 | 7.33 | 三间房镇、常营、管庄3个街道 | |
合作网点 | 1/2 | 113 | 25.53 | 1.92 | 建外、六里屯、双井、呼家楼、平房镇等23个街道 |
2/5 | 61 | 18.39 | 1.58 | 学院路、奥运村、北下关、紫竹院等10个街道 | |
3/0 | 35 | 12.72 | 2.06 | 东铁匠营、方庄、十八里店、潘家园等8个街道 | |
4/4 | 33 | 12.96 | 1.92 | 德胜、和平街、安贞、和平里、太阳宫等8个街道 | |
5/3 | 30 | 13.33 | 1.87 | 西长安街、展览路、新街口、广安门内等5个街道 |
表4 北京市路网密度与顺丰快递网点密度Tab. 4 Beijing road network density and Shunfeng express service outlet density |
县区 | 网点密度/(个/km2) | 路网密度/(km/km2) |
---|---|---|
海淀区 | 1.80 | 7.89 |
朝阳区 | 2.40 | 8.16 |
西城区 | 4.18 | 16.87 |
东城区 | 3.35 | 17.91 |
丰台区 | 1.96 | 6.34 |
石景山区 | 1.54 | 7.41 |
表5 北京市各区各类公共服务设施占比情况Tab. 5 Proportions of various public service facilities in the urban districts of Beijing |
海淀区/% | 朝阳区/% | 西城区/% | 东城区/% | 丰台区/% | 石景山区/% | 总数/个 | |
---|---|---|---|---|---|---|---|
政府机关单位 | 32.75 | 16.08 | 15.80 | 14.61 | 15.40 | 5.36 | 7407 |
教育 | 38.17 | 29.68 | 10.63 | 7.63 | 10.60 | 3.29 | 4063 |
金融 | 29.25 | 30.69 | 13.77 | 9.74 | 13.91 | 2.64 | 2229 |
医院 | 23.20 | 35.14 | 11.80 | 11.26 | 14.38 | 4.22 | 737 |
[1] |
中华人民共和国国家国家邮政局. 快递业发展“十三五”规划[EB/OL]. http://www.askci.com/news/chanye/20170 215/14105590554.shtml.
[ State Post Bureau of The People's Republic of China. The 13th five-year plan for the development of express delivery industry[EB/OL]. http://www.askci.com/news/chanye/20170215/14105590554.shtml.]
|
[2] |
中华人民共和国商务部. 中国电子商务报告2018[R]. 北京: 2019 -05-29.
[ Ministry of Commerce of The People's Republic of China. China e-commerce report 2018[R]. Beijing:2019-05-29. ]
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
倪玲霖, 王姣娥, 胡浩. 中国快递企业的空间组织研究——以顺丰速运为例[J]. 经济地理, 2012,32(2):82-88.
[
|
[9] |
李鲁奇, 马学广. 基于运单数据的中国快递型物流网络格局、结构与功能——以宅急送为例[J]. 地理科学, 2019,39(1):92-100.
[
|
[10] |
李波. 快递企业服务网络布局优化研究[D]. 北京:北京交通大学, 2010.
[
|
[11] |
谢夏成. 上海地区快递物流节点的空间格局分析[D]. 上海:上海师范大学, 2015.
[
|
[12] |
黄涛. 基于GIS的快递末端自提网点布局研究[D]. 西安:长安大学, 2017.
[
|
[13] |
李妮娜. 基于中心地理论的城市快递服务网点选址研究[D]. 北京:北京交通大学, 2015.
[
|
[14] |
谭如诗, 徐逸伦, 陈栋. 城市居民快递自提行为空间研究——以南京市城区菜鸟驿站为例[J]. 世界地理研究, 2016,25(5):111-120.
[
|
[15] |
张珣, 陈健璋, 黄金川. 基于空间聚类方法的京津冀城市群多层级空间结构研究[J]. 地理科学进展, 2017,36(11):41-49.
[
|
[16] |
湛东升, 张文忠, 党云晓. 北京市公共服务设施空间集聚特征分析[J]. 经济地理, 2018,38(12):78-84.
[
|
[17] |
宋辞, 裴韬. 北京市多尺度中心特征识别与群聚模式发现[J]. 地球信息科学学报, 2019,21(3):84-97.
[
|
[18] |
浩飞龙. 多中心视角下的长春市城市功能空间结构研究[D]. 长春:东北师范大学, 2017.
[
|
[19] |
王劲峰, 廖一兰, 刘鑫. 空间数据分析教程[M]. 北京: 科学出版社, 2010.
[
|
[20] |
|
[21] |
杨帆, 徐建刚, 周亮. 基于DBSCAN空间聚类的广州市区餐饮集群识别及空间特征分析[J]. 经济地理, 2016,36(10):110-116.
[
|
[22] |
申庆喜, 李诚固, 刘仲仪. 长春市公共服务设施空间与居住空间格局特征[J]. 地理研究, 2018,37(11):127-136.
[
|
[23] |
|
[24] |
李钢, 陈未雨, 杨兰. 武汉市快递自提点的空间格局与集聚模式研究[J]. 地理科学进展, 2019,38(3):105-114.
[
|
[25] |
李钢, 杨兰, 贺建雄. 基于POI数据的西安市快递自提点空间格局及空间关系研究——以菜鸟驿站为例[J]. 地理科学, 2018,38(12):85-91.
[
|
[26] |
李江苏, 梁燕, 王晓蕊. 基于POI数据的郑东新区服务业空间聚类研究[J]. 地理研究, 2018,37(1):147-159.
[
|
[27] |
刘玲, 李钢, 杨兰, 等. 深圳市快递自提点的空间分布特征与影响因素[J]. 地球信息科学学报, 2019,21(8):1240-1253.
[
|
[28] |
国家科学技术部. 中国统计年鉴(2017)[M]. 北京: 中国统计出版社, 2017.
[ Ministry of Science and Technology of the People's Republic of China. Chinese science and technology statistics yearbook-2017. Beijing: China Statistics Press, 2017. ]
|
/
〈 | 〉 |