利用可达性模型预测轨道交通站点发生客流

Predicting Ridership at Rail Transit Station with Accessibility Model

该文探讨了基于站点可达性和周边用地的轨道站点发生客流预测方法。在GIS中,通过双重约束蒙特卡罗模拟算法将居住人口分解到小尺度栅格上。根据可达性中的距离衰减原理,采用罗切斯特模型计算居民选择轨道交通的概率。基于各栅格位置的人口数量及其与轨道站点距离获得的出行概率,获得轨道站点的客流发生量。距离衰减函数的参数配置决定了客流发生预测的质量。以武汉市为例,对参数配置进行了探讨。与实际调查数据验证表明,通过选取适当的参数,基于GIS可达性的预测模型可以达到较好的精度。

We explore a forecasting method of trip production at rail transit station based on the accessibility. The accessibility model makes use of land use and population distribution data. In GIS, population data from census zones may be disaggregated into smaller raster cells. In accessibility model, negative logistic function based on distance decay is applied. Transit trip production at each rail station may be achieved by combining population locations with the accessibility model. During the process, parameters of the accessibility model play a key role. The parameters are tested in the case of Wuhan city. Through proper parameter configuration, the accessibility model may achieve acceptable accuracy for predicting rail transit trip production.