Acta Scientiarum Naturalium Universitatis Pekinensis ›› 2022, Vol. 58 ›› Issue (4): 664-672.DOI: 10.13209/j.0479-8023.2022.060

Previous Articles     Next Articles

Analysis of Urban Land Use Function Identification in Shenzhen Based on SOFM Network

WU Jianan1,2, CHU Jun1,2, SUN Yiyu1,2, CHAO Heng2,3,†   

  1. 1. School of Urban Planning and Design, Peking University Shenzhen Graduate School, Shenzhen 518055 2. Laboratory for Urban Future, Peking University Shenzhen Graduate School, Shenzhen 518055 3. Shenzhen New Land Tool Planning &Architectural Design Co., Ltd, Shenzhen 518055
  • Received:2021-07-17 Revised:2021-08-16 Online:2022-07-20 Published:2022-07-20
  • Contact: CHAO Heng, E-mail: chaoheng_0827(at)126.com

SOFM网络下的深圳市城市用地功能识别分析

吴佳楠1,2, 储君1,2, 孙裔煜1,2, 晁恒2,3,†    

  1. 1. 北京大学深圳研究生院城市规划与设计学院, 深圳 518055 2. 北京大学(深圳)未来城市实验室, 深圳 518055 3. 深圳市新城市规划建筑设计股份有限公司, 深圳 518055
  • 通讯作者: 晁恒, E-mail: chaoheng_0827(at)126.com
  • 基金资助:
    国家自然科学基金(42071204)、广东省基础与应用基础研究基金(2019A1515011661)和北京大学未来城市实验室(深圳)铁汉科研开放课题基金项目资助

Abstract:

Based on 88814 POI data and with the help of Self-Organizing Feature Map (SOFM), the functions of urban land use in Shenzhen are analyzed and identified, and compared with China’s basic urban land use type mapping (EULUC-China). It is found that 1) the urban block scale of Shenzhen presents the characteristics of mixed functions, which is mainly divided into five types of spatial land, such as secondary industry and producer service industry, transportation facilities, residence + secondary industry and producer service industry + commercial service industry facilities, residence + public management and service, commerce + public management and service. 2) The land use function identification model based on SOFM network has more detailed results, and the classification of mixed function land can highlight the regional characteristics.

Key words: POI, SOFM, land use, urban functional area identification, Shenzhen city

摘要:

基于88814条POI数据, 并借助自组织映射网络(SOFM网络), 对深圳市城市土地利用的功能进行分析和识别, 并与中国基本城市土地利用类型制图(EULUC-China)进行比较验证。结果表明: 1) 深圳市城市街区尺度呈现功能混合的特征, 主要分为第二产业和生产性服务业主导、交通设施主导、居住+第二产业及生产性服务业+商业服务业设施混合、居住+公共管理和服务混合以及商业+公共管理和服务混合5类空间用地。2) 基于SOFM网络的用地功能识别模型, 结果更为细致, 混合功能用地的分类更能突出地域特征。

关键词: 兴趣点(POI), 自组织映射网络(SOFM)网络, 土地利用, 功能识别, 深圳市