K-shape-based spatio-temporal graph attention network for traffic flow prediction

Jin Zhang; Yimin Yang; Xiaoheng Wu; Honggang Wang

doi:10.1117/12.3050703

20 January 2025 K-shape-based spatio-temporal graph attention network for traffic flow prediction

Jin Zhang, Yimin Yang, Xiaoheng Wu, Honggang Wang

Author Affiliations +

Proceedings Volume 13422, Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024); 134221N (2025) https://doi.org/10.1117/12.3050703
Event: Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024), 2024, Xi'an, China

Abstract

Traffic flow prediction is a crucial component of intelligent transportation systems, providing scientific support for traffic management and urban planning. However, current studies are most focused on short-term prediction, with long-term prediction remaining a challenge. In addition, to control the dynamic spatio-temporal characteristics of urban traffic effectively is still difficult for traditional methods. To address these challenges, this paper proposes a K-shape-based spatiotemporal graph attention network model (KSTGAT) for long-term traffic flow prediction. The model employs a multi-head attention mechanism to extract long-term temporal features and replaces the traditional linear mappings in the attention mechanism with GRU to better capture nonlinear dynamic changes. For spatial feature extraction, K-shape is used to cluster the input sequences, and a composite adjacency matrix is constructed based on the clustering results and predefined adjacency matrix. This is followed by a graph attention network to extract dynamic spatial features. Additionally, the model also incorporates the periodic characteristics of traffic flow data. Experimental results demonstrate that KSTGAT outperforms state-of-the-art models, proving its effectiveness in long-term prediction.

(2025) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Jin Zhang, Yimin Yang, Xiaoheng Wu, and Honggang Wang "K-shape-based spatio-temporal graph attention network for traffic flow prediction", Proc. SPIE 13422, Fourth International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2024), 134221N (20 January 2025); https://doi.org/10.1117/12.3050703

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

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
7 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Matrices

Feature extraction

Data modeling

Machine learning

Semantics

Performance modeling

Convolution

Show All Keywords

Keywords/Phrases

Search In:

Publication Years