Identifying and predicting framework of urban agglomeration air-rail intermodal passenger flow based on signaling data
CHEN Yanyan
ZHANG Ye
ZHANG Yunchao
LI Yongxing
LI Chen
LAI Jianhui
Abstract：To address the challenges of obtaining demand for air-rail passenger flows within urban agglomera-tions and understanding passenger flow patterns,a comprehensive analysis framework integrating air-railway passenger flow identification and prediction modules is proposed. Firstly,considering the spatial constraints of transportation hubs and the spatiotemporal characteristics of travel,a method for identifying intermodal pas-senger flows using signaling data is introduced,accompanied by an analysis of distribution patterns. Subse-quently,leveraging the Bidirectional Gated Recurrent Unit (BiGRU),time period coding is integrated to construct a Temporal-Bidirectional Gated Recurrent Unit (T-BiGRU) for passenger flow prediction. Finally,the framework is validated using the Beijing-Tianjin-Hebei urban agglomeration as a case study. Results indi-cate that the intermodal passenger flow in the Beijing-Tianjin-Hebei urban agglomeration exhibits a clustered distribution,with the scenarios of Beijing South Railway Station-Tianjin Railway Station-Tianjin Binhai Air-port and Beijing West Railway Station-Zhengding Airport Station-Shijiazhuang Zhengding Airport having the highest proportion,exceeding 65％ of the total intermodal passenger flow. The T-BiGRU model accurately predicts the demand for air-rail intermodal passenger flows. The bidirectional passenger flow prediction accuracy for the two main scenarios exceeds 89％,surpassing multiple baseline models. These findings offer support for the coordinated development of air-rail transportation and the optimization of air-rail intermodal services in urban agglomerations.
Keywords：integrated transportationair-rail intermodalpassenger flow identificationT-BiGRU modelurban agglomeration
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( 1-10 )