基于精细化用地的轨道客流直接估计模型

现有轨道客流直接估计模型中，对于用地的描述多基于人口、岗位分布、用地面积等概略数据，未能准确反映用地开发的多样性和复杂性，难以揭示用地混合开发对于居民出行的作用机理.本文采用北京市电子地图中的兴趣点（POI）数据表示用地信息，结合轨道站点多层次吸引范围划分，实现了轨道站点吸引范围内各类用地比例的精细化描述.通过将精细化用地信息与轨道站点乘降客流量进行回归分析，总结了用地、交通、区位因素对于轨道站点乘降客流的影响机理，建立了基于精细化用地的轨道站点客流估计模型.验证结果显示，模型对于本文所研究典型站点的预测精度达到预期，反映了用地与轨道客流之间的强相关性.

Station Level Transit Ridership Direct Estimation Model Based on Precise Land Use

The existing transit ridership models describe land use with indexes such as population, employment and land use area, these indexes can' t describe the diversity and complexity of land use development, thus make it difficult to reveal the influence mechanism of mixed land use on travel demand. This paper quantifies the precise land use characteristics of Beijing transit station catchment areas using POIs (points of interest) extract from an electronic map. Through the regression analysis of precise land use and transit station boarding volume, a direct ridership model is built based on the influence mechanism analysis of land use, traffic and location factors on transit ridership. Following validation shows the prediction accuracy of the proposed model meets expectation, reveals the strong correlation between land use and transit ridership.