利用出租车轨迹数据识别城市功能区

Identifying City Functional Areas Using Taxi Trajectory Data

城市系统的复杂性使得利用出租车轨迹数据理解城市的功能区结构存在结构分区不合理、识别结果精度低及不可靠的问题。对于城市功能区研究单元划分问题,从出租车轨迹空间分布特征入手,提出了一种基于限制条件和轨迹中心线包裹的城市区域划分方法。对于功能区理解问题,针对娱乐地、工作地和居住地的出租车OD(上车点和下车点)时间分布特征,提出从娱乐区、居住区到工作区逐步进阶的功能区识别方法。分别构造了娱乐因子和居住因子对娱乐区和居住区进行提取;结合工作地出租车下车点时间分布特征,利用监督分类对工作地进行了提取。

Understanding and analyzing city functional area by taxi trajectory data has the problems of unreasonable structural division, low accuracy and unreliable identification results because of the complexity of urban system.Firstly, an urban area subdivision method based on constraints and trajectory centerline wrapping is proposed to solve the problem of urban functional area division. Secondly, the characteristics of OD temporal distribution in entertainment areas, job areas and residential areas are explored respectively and a functional area identification method that gradually advances from entertainment area, residential area to the working area is put forward to understand functional area. Finally, entertainment factor and residential factor are respectively constructed to extract entertainment area and residential area. And the working area is extracted using supervised classification method which combines with the time distribution characteristics of the taxi drop-off point at the working area.