Simulation and optimization of pedestrian transfer flow in high-speed railway hubs and urban rail transit systems
CHENG Long
NING Zhe
XUE Xiaoyu
ZHANG Jiyang
LIU Zhipeng
Abstract：This study addresses the issues related to the optimization of transfer flows between high-speed rail hubs and urban rail transit,focusing on Nanjing South Railway Station as the case study. A full-process simulation model is developed to identify transfer bottlenecks,using the Gradient Boosting Decision Tree (GBDT) to determine parameter importance and propose improvement measures. Firstly,the pedestrian and facility flows of the transfer process are decomposed to analyze the distribution characteristics of passenger flows from high-speed rail to urban rail transit. Secondly,AnyLogic is employed to create a full-process simulation model of the transfer at high-speed rail hubs,analyzing the current simulation results to identify spatial bottlenecks. Then various optimization measures are designed under different types,adjusting parameters to formulate combination schemes. The GBDT algorithm is used to ascertain the relative importance of different measures and parameters,establishing the priority of improvement measures. Finally,based on this prioritization,different types of optimization combination schemes are determined,with machine learning interpretability methods applied to analyze their effectiveness,providing recommendations for service enhancements across different scenarios. Results show that transfer bottlenecks are primarily located at stair/escalator passageways and service facilities such as gates and ticket machines. The proportion of urban rail ticket purchases significantly influences both average transfer time and the maximum number of transfers per unit time. For average transfer time,the number of automatic ticket machines,ticket purchase time,urban rail entry service time,and the number of entry gates have a relatively large impact. For maximum transfers per unit time,ticket purchase time and entry gate service time are most influential. To improve transfer efficiency at high-speed railway hubs,it is recommended to promote electronic tickets and various payment methods,and optimize ticketing and checking facilities.
Keywords：multimodal transportation systemtransportation hubfull-process transfer optimizationAnyLogic simulationGradient Boosting Decision Tree (GBDT)
Publication Date：2024-08-28
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：10( 43-52 )