Short-term passenger flow forecast of urban rail transit based on GAN

Hua Li; Chuang Zhu; Haoran Li

doi:10.1117/12.2668732

16 February 2023 Short-term passenger flow forecast of urban rail transit based on GAN

Hua Li, Chuang Zhu, Haoran Li

Proceedings Volume 12591, Sixth International Conference on Traffic Engineering and Transportation System (ICTETS 2022); 125913B (2023) https://doi.org/10.1117/12.2668732
Event: 6th International Conference on Traffic Engineering and Transportation System (ICTETS 2022), 2022, Guangzhou, China

Abstract

Short-term passenger flow forecasting plays an important role in better traffic management and organization. Emerging deep learning models offer a great way to improve short-term forecast accuracy. However, a large number of existing forecasting models are very complex in their model structure because they combine various neural network layers to improve the accuracy of the models. Therefore, it is necessary to reduce the complexity of the models while ensuring their accuracy. Finally, we propose a Graph-GAN model based on deep learning with a simple structure and high prediction accuracy to predict short-term passenger flows in urban rail networks. Graph-GAN consists of two main components: (1) a graph convolutional network (GCN) for extracting topological information of the metro network; and (2) a generative adversarial network (GAN) for predicting passenger flow, where the generator and discriminator in the GAN consist of only simple fully connected neural networks. Graph-GAN was tested on real Beijing subway datasets in 2016 and 2018. The prediction performance of Graph-GAN is compared with several state-of-the-art models to illustrate the superiority of the model.

Citation Download Citation

Hua Li, Chuang Zhu, and Haoran Li "Short-term passenger flow forecast of urban rail transit based on GAN", Proc. SPIE 12591, Sixth International Conference on Traffic Engineering and Transportation System (ICTETS 2022), 125913B (16 February 2023); https://doi.org/10.1117/12.2668732

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available