ABSTRACT Metro capacity should be further improved with growing demand for metros. For metro lines with high-frequency services, the train dwell time is a determinant of the number of trains per hour. Uneven passenger distribution on the platform may cause a longer dwell time, thereby reducing the capacity. This study analyses the mechanism of the passenger distribution on the platform with regard to which cars of a train passengers choose to board. A model has been developed, considering the origin-destination and waiting time of passengers and the platform layouts of the stations where passengers board and alight. The model was calibrated using loadweigh data, which provides the weight of passengers on each car of a train, on the Hammersmith & City line of the London Underground. The largest difference between the actual and modelled average number of boarders on a car is less than 2 passengers, which suggests the model performs well. The results show, at both stations studied, 44% of passengers in the morning peak chose boarding cars based on the layouts of the stations where they alighted. The results would be useful for metro operation planners and station staff to take measures to manage passenger distribution on the platform.

中文翻译：

---

基于乘客上车选择的地铁站台乘客分布建模

摘要 随着地铁需求的增长，地铁容量应进一步提高。对于提供高频服务的地铁线路，列车停留时间是每小时列车数量的决定因素。平台上的乘客分布不均可能会导致停留时间更长，从而降低容量。本研究分析了列车乘客选择上车的站台乘客分布机制。综合考虑旅客始发地、候车时间和旅客上下车站站台布局，建立模型。该模型使用负载重量数据进行校准，该数据提供了伦敦地铁 Hammersmith & City 线上火车每节车厢乘客的重量。汽车上实际和模型的平均寄宿生人数之间的最大差异小于 2 名乘客，这表明该模型性能良好。结果显示，在所研究的两个站点中，44% 的早高峰乘客根据下车站点的布局选择了上车。结果将有助于地铁运营规划人员和车站工作人员采取措施管理平台上的乘客分布。