地理科学进展 ›› 2018, Vol. 37 ›› Issue (8): 1106-1118.doi: 10.18306/dlkxjz.2018.08.010

• 研究论文 • 上一篇    下一篇

基于居民行为周期特征的城市空间研究

钟炜菁1(), 王德2,*()   

  1. 1. 杭州市城市规划设计研究院,杭州 310012
    2. 同济大学建筑与城市规划学院,上海 200092
  • 收稿日期:2017-11-03 修回日期:2018-01-07 出版日期:2018-09-04 发布日期:2018-09-04
  • 通讯作者: 王德 E-mail:wjzhong0705@126.com;dewang@tongji.edu.cn
  • 作者简介:

    作者简介:钟炜菁(1991- ),女,江西赣州人,硕士,规划师,研究方向为城市空间与行为、数据分析,E-mail: wjzhong0705@126.com

  • 基金资助:
    国家自然科学基金项目(41771170)

Urban space study based on the temporal characteristics of residents' behavior

Weijing ZHONG1(), De WANG2,*()   

  1. 1. Hangzhou City Planning and Design Academy, Hangzhou 310012, China
    2. College of Architecture and Urban Planning Tongji University, Shanghai 200092, China
  • Received:2017-11-03 Revised:2018-01-07 Online:2018-09-04 Published:2018-09-04
  • Contact: De WANG E-mail:wjzhong0705@126.com;dewang@tongji.edu.cn
  • Supported by:
    National Natural Science Foundation of China, No. 41771170

摘要:

伴随着中国经济社会进入“新常态”的发展阶段,对城市存量空间的研究提出了更加精细化的要求,基于居民行为活动的周期规律对城市空间进行研究,进而提升城市空间的品质日益重要。随着信息通信技术的快速发展,使许多大数据的获取成为可能,并由于其低成本、即时、大样本等优势,在城市空间研究方面具有巨大的价值。以上海市中心城区为例,利用手机信令数据,探究居民活动的空间周期变化特征,并基于空间的周期特征曲线,采用相似性传播聚类算法进行空间分类。研究表明,居民活动有平日一日周期和平日加周末二日周期,与人的作息规律相符合。市核心区、城市副中心及主要就业中心,昼夜波动和平日周末活动强度的差异都较为明显。空间分类结果显示,城市活动空间的组织既体现出个体充分的空间能动性,也反映出对土地使用类型以及设施建设、投入程度的耦合性。上海市内环内核心区混合多样的用地模式使得活动区内居民活动内容丰富,周期特征功能区边界模糊。研究成果可为未来的城市空间规划提供指导,为城市空间结构、功能布置、设施布局等优化提供决策支撑和科学依据。

关键词: 手机信令数据, 周期特征, 居民行为, 空间分类, 近邻相似性传播聚类, 上海

Abstract:

As the development of economy and society enters into the "new normal" stage in China, urban planning is also gradually transformed from the traditional incremental planning to inventory planning. It is important to explore the urban spatial dynamic functional characteristics, and to optimize the use of urban activity space based on people's needs, which would enhance the quality of urban space. Advancements of information, communication, and location-aware technologies have made collections of various passively generated datasets possible. These datasets provide new opportunities to understand spatial dynamic characteristics at a low cost and large scale. This study explored the classification of urban space and spatial dynamic characteristics based on a large mobile phone location dataset from Shanghai Municipality, China. The results suggest that the geographical differences of spatial dynamic patterns in Shanghai are evident. The diurnal activity curve is consistent with the patterns of human activity. There were significant differences in intensity of day-to-day activity fluctuations and weekday activities between downtown, sub-centers, and major employment centers. Affinity propagation clustering was introduced to identify the characteristics of urban spatial structure and identify the characteristics of urban space structure of liquidity and viscosity. Several distinct patterns were extracted, and the spatial distributions of the derived clusters highlight distinct human mobility patterns in different areas of the city. We then discuss the socioeconomic and demographic characteristics of the regions covered by different cluster types to gain insights of human mobility patterns in the context of urban functional regions. The findings could offer useful information for policy and decision making.

Key words: mobile phone signaling data, temporal characteristics, residents' behavior, spatial clustering, Affinity Propagation Clustering(AP), Shanghai