基于居民出行特征的北京城市功能区识别与空间交互研究

受区域功能分化影响,城市居民出行呈现出特定的时序特征,因而不同的出行时序特征可以反映区域功能的差异性。同时,区域功能的交互特征可以通过居民出行的空间交互活动体现。大数据时代的到来,使得以GPS数据为代表的个体时空大数据可以从微观视角反映居民出行特征。本文采用个体时空大数据,应用数据挖掘方法,从居民感知视角研究城市区域功能的差异性与联系性。以北京六环为研究区域,采用规则格网划分城市地块,通过北京市3个月的出租车GPS数据提取地块的居民出行时序特征。采用期望最大化算法进行聚类分析,并结合兴趣点数据和居民出行调查实现功能区识别,识别出居住区、商业娱乐区等6类功能区。从距离和时间2个维度分析功能区之间的空间交互特征,发现功能互补性在一定程度上削弱了空间交互强度的距离衰减效应,同时功能交互呈现出显著的时序差异。

Identification and Spatial Interaction of Urban Functional Regions in Beijing Based on the Characteristics of Residents' Traveling

Affected by the differentiation of regional functions, the urban residents' activities presents specific timing characteristics. Different traveling patterns could indicate differences in regional functions. Meanwhile, the interactive features of regional function can be reflected by the spatial interaction activities of residents' trips. The advent of the big data era makes the individual geographical big data represented by the GPS data feasible to reflect residents' trip characteristics from the micro perspective. In this paper, the individual geographical big data and data mining method are employed to study the diversities and connections of urban regional functions under the perspective of residential perception. The study area enclosed by the Sixth Ring Road in Beijing is divided into regular grids, for the convenience of extracting timing characteristics of residential activities from 3 months' GPS data on taxis. Specifically, the cluster analysis based on expectation maximization algorithm, the point of interest and the daily traveling characteristics of residents are used to identify functional regions into six types, such as residential districts and commercial entertainment districts. Finally, the spatial interaction characteristics between functional areas are analyzed from two dimensions of distance and time, revealing that functional complementation weakens the influence of distance on the spatial interaction strength and the functional interaction indicates significant temporal differences.