Multi-View Spatio-Temporal Dynamic Graph Convolution Network for Traffic Flow Prediction

• Published in: IEEE internet of things journal p. 1
• Main Authors: Zhong, Lihong, Wang, Bin, Tian, Zhao, Rodrigues, Tiago Koketsu, Liu, Wei, She, Wei
• Format: Journal Article
• Language: English
• Published: IEEE 2025
• Subjects: Adaptation models; Convolution; Correlation; Feature extraction; graph neural network; Graph neural networks; Intelligent transportation system; multi-view learning; Predictive models; Roads; spatio-temporal dependence; traffic flow prediction; Transportation; Vehicle dynamics
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2025.3580788

Accurate traffic flow prediction is crucial for alleviating traffic congestion and optimizing intelligent transportation systems. However, traffic flow is subject to uncertainties and exhibits complex spatial and temporal dependence and dynamic change characteristics. Moreover, many efforts rely on a single view, which makes it difficult to comprehensively capture multiple levels of spatial and temporal correlations, thus limiting the accuracy of predictions. Therefore, we propose the multi-view spatio-temporal dynamic graph convolution framework MVSTDG for more comprehensively exploring and fusing the multi-view spatio-temporal features. Firstly, we design a dual-path Time-Patch Convolution (TPConv) module to separately model short-term fluctuations and long-term periodic trends, enabling effective extraction of dynamic features at multiple temporal scales. Secondly, we construct a data-driven traffic pattern library to generate dynamic adjacency matrices and integrate them with static topologies view. An Adaptive Diffusion Graph Convolutional Network (ADGCN) is then employed to model both local and global spatial correlations. In addition, we design a cross-gated spatio-temporal fusion mechanism that adaptively adjusts the contribution of short-term and long-term information, enhances the interaction of spatio-temporal information, and improves the model's adaptive capability under different time scales. The experimental results show that MVSTDG outperforms the state-of-the-art baselines in several evaluation metrics and demonstrates higher prediction accuracy and stability on the four real datasets.