Forecasting the subway passenger flow under event occurrences with multivariate disturbances

作者：

Highlights：

• Multivariate disturbances are considered to forecast the passenger flow, and a hybrid deep learning framework is proposed.

• Constructing an inbound matrix from other stations of the target station can promote the performance of the model.

• The tendency of audience can be extracted from social media which promotes the performance of the model.

摘要

•Multivariate disturbances are considered to forecast the passenger flow, and a hybrid deep learning framework is proposed.•Constructing an inbound matrix from other stations of the target station can promote the performance of the model.•The tendency of audience can be extracted from social media which promotes the performance of the model.

论文关键词：Social media,Subway passenger flow prediction,Attention,Spatiotemporal disturbances

论文评审过程：Received 19 June 2020, Revised 3 April 2021, Accepted 8 October 2021, Available online 18 October 2021, Version of Record 20 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116057