Spatio-temporal graph data have been widely applied in traffic flow prediction tasks. Traditional methods often combine graph convolutional networks with recurrent neural networks based on the original graph structure. However, these methods typically struggle with issues such as incomplete sensor distributions and unresolved potential dependencies between different parallel intersections. Moreover, the model structure often fails to capture the noise inherent in traffic predictions, which becomes a significant barrier to accurate forecasting. To address these challenges, we proposed a novel approach that diverges from traditional traffic flow models, which rely on predefined graph structures. Instead, our method uncovers hidden edge connectivity and integrates the data into a unified framework, enabling the model to better capture the underlying spatial relationships. Specifically, we introduced a new framework that combines neural control differential equations with stochastic differential equations to enhance spatio-temporal graph modeling. This integration improves the model's capacity to capture complex spatio-temporal patterns, thereby enhancing the accuracy of prediction tasks. Extensive experiments conducted on benchmark datasets validate the effectiveness of our method.
Citation: Ruxin Xue, Jinggui Huang, Zaitang Huang, Bingyan Li. Reconstructed graph spatio-temporal stochastic controlled differential equation for traffic flow forecasting[J]. Electronic Research Archive, 2025, 33(4): 2543-2566. doi: 10.3934/era.2025113
Spatio-temporal graph data have been widely applied in traffic flow prediction tasks. Traditional methods often combine graph convolutional networks with recurrent neural networks based on the original graph structure. However, these methods typically struggle with issues such as incomplete sensor distributions and unresolved potential dependencies between different parallel intersections. Moreover, the model structure often fails to capture the noise inherent in traffic predictions, which becomes a significant barrier to accurate forecasting. To address these challenges, we proposed a novel approach that diverges from traditional traffic flow models, which rely on predefined graph structures. Instead, our method uncovers hidden edge connectivity and integrates the data into a unified framework, enabling the model to better capture the underlying spatial relationships. Specifically, we introduced a new framework that combines neural control differential equations with stochastic differential equations to enhance spatio-temporal graph modeling. This integration improves the model's capacity to capture complex spatio-temporal patterns, thereby enhancing the accuracy of prediction tasks. Extensive experiments conducted on benchmark datasets validate the effectiveness of our method.
| [1] |
T. Zhang, W. Ding, T. Chen, Z. Wang, J. Chen, A graph convolutional method for traffic flow prediction in highway network, Wireless Commun. Mobile Comput., 2021 (2021), 1997212. https://doi.org/10.1155/2021/1997212 doi: 10.1155/2021/1997212
|
| [2] | B. N. Oreshkin, A. Amini, L. Coyle, M. Coates, FC-GAGA: Fully connected gated graph architecture for spatio-temporal traffic forecasting, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 9233–9241. https://doi.org/10.1609/aaai.v35i10.17114 |
| [3] | X. Zhang, C. Huang, Y. Xu, L. Xia, P. Dai, L. Bo, et al., Traffic flow forecasting with spatial-temporal graph diffusion network, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 15008–15015. https://doi.org/10.1609/aaai.v35i17.17761 |
| [4] |
M. A. Zaytar, C. E. Amrani, Sequence to sequence weather forecasting with long short-term memory recurrent neural networks, Int. J. Comput. Appl., 143 (2016), 7–11. https://doi.org/10.5120/ijca2016910497 doi: 10.5120/ijca2016910497
|
| [5] | S. F. Tekin, O. Karaahmetoglu, F. Ilhan, I. Balaban, S. S. Kozat, Spatio-temporal weather forecasting and attention mechanism on convolutional lstms, preprint, arXiv: 2102.00696. |
| [6] | M. Hossain, B. Rekabdar, S. J. Louis, S. Dascalu, Forecasting the weather of Nevada: a deep learning approach, in 2015 International Joint Conference on Neural Networks (IJCNN), Killarney, Ireland, (2015), 1–6. https://doi.org/10.1109/IJCNN.2015.7280812 |
| [7] | T. Alghamdi, K. Elgazzar, M. Bayoumi, T. Sharaf, S. Shah, Forecasting traffic congestion using ARIMA modeling, in 2019 15th International Wireless Communications & Mobile Computing Conference (IWCMC), Tangier, Morocco, (2019), 1227–1232. https://doi.org/10.1109/IWCMC.2019.8766698 |
| [8] |
Y. Yılmaz, Machine learning-enhanced traffic light optimization system prioritizing emergency vehicle passage using SVM and random forest models, Gazi Univ. J. Sci. Part A: Eng. Innov., 12 (2025), 175–196. https://doi.org/10.54287/gujsa.1581105 doi: 10.54287/gujsa.1581105
|
| [9] | B. Yu, H. Yin, Z. Zhu, Spatio-temporal graph convolutional networks: a deep learning framework for traffic forecasting, preprint, arXiv: 1709.04875. https://doi.org/10.24963/ijcai.2018/505 |
| [10] | G. Chen, L. Hu, Q. Zhang, Z. Ren, X. Gao, J. Cheng, ST-LSTM: spatio-temporal graph based long short-term memory network for vehicle trajectory prediction, in 2020 IEEE International Conference on Image Processing (ICIP), (2020), 608–612. https://doi.org/10.1109/ICIP40778.2020.9191332 |
| [11] | Z. Qiu, K. Qiu, J. Fu, D. Fu, Dgcn: dynamic graph convolutional network for efficient multi-person pose estimation, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 11924–11931. https://doi.org/10.1609/aaai.v34i07.6867 |
| [12] |
S. Yang, Q. Wu, Y. Wang, Z. Zhou, MSTDFGRN: a multi-view spatio-temporal dynamic fusion graph recurrent network for traffic flow prediction, Comput. Electr. Eng., 123 (2025), 110046. https://doi.org/10.1016/j.compeleceng.2024.110046 doi: 10.1016/j.compeleceng.2024.110046
|
| [13] |
S. Yang, Q. Wu, Y. Wang, T. Lin, SSGCRTN: a space-specific graph convolutional recurrent transformer network for traffic prediction, Appl. Intell., 54 (2024), 11978–11994. https://doi.org/10.1007/s10489-024-05815-1 doi: 10.1007/s10489-024-05815-1
|
| [14] |
S. Yang, Q. Wu, SDSINet: A spatiotemporal dual-scale interaction network for traffic prediction, Appl. Soft Comput., 173 (2025), 112892. https://doi.org/10.1016/j.asoc.2025.112892 doi: 10.1016/j.asoc.2025.112892
|
| [15] | S. Yang, Q. Wu, Z. Li, K. Wang, PSTCGCN: principal spatio-temporal causal graph convolutional network for traffic flow prediction, Neural Comput. Appl., 2024 (2024). https://doi.org/10.1007/s00521-024-10591-7 |
| [16] |
Z. Mei, X. Bi, D. Li, W. Xia, F. Yang, H. Wu, DHHNN: a dynamic hypergraph hyperbolic neural network based on variational autoencoder for multimodal data integration and node classification, Inf. Fusion, 119 (2025), 103016. https://doi.org/10.1016/j.inffus.2025.103016 doi: 10.1016/j.inffus.2025.103016
|
| [17] |
L. Hu, L. Wei, Y. Lin, Decomposition dynamic multi-graph convolutional recurrent network for traffic forecasting, Appl. Intell., 55 (2025), 1–17. https://doi.org/10.1007/s10489-025-06503-4 doi: 10.1007/s10489-025-06503-4
|
| [18] | C. Song, Y. Lin, S. Guo, H. Wan, Spatial-temporal synchronous graph convolutional networks: a new framework for spatial-temporal network data forecasting, in Proceedings of the AAAI Conference on Artificial Intelligence, 34 (2020), 914–921. https://doi.org/10.1609/aaai.v34i01.5438 |
| [19] | S. Guo, Y. Lin, N. Feng, C. Song, H. Wan, Attention based spatial-temporal graph convolutional networks for traffic flow forecasting, in Proceedings of the AAAI Conference on Artificial Intelligence, 33 (2019), 922–929. https://doi.org/10.1609/aaai.v33i01.3301922 |
| [20] | H. He, K. Ye, Graph structure neural differential equations on spatio-temporal prediction, in 2022 IEEE International Conference on Big Data (Big Data), (2022), 1830–1835. https://doi.org/10.1109/BigData55660.2022.10020863 |
| [21] | X. Zheng, Y. Wang, Y. Liu, M. Li, M. Zhang, D. Jin, et al., Graph neural networks for graphs with heterophily: a survey, preprint, arXiv: 2202.07082. https://doi.org/10.48550/arXiv.2202.07082 |
| [22] | J. Zhu, R. A. Rossi, A. Rao, T. Mai, N. Lipka, N. K. Ahmed, et al., Graph neural networks with heterophily, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 11168–11176. https://doi.org/10.1609/aaai.v35i12.17332 |
| [23] | R. K. Yadav, A. Abhishek, S. Sourav, S. Verma, GCN with clustering coefficients and attention module, in 2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE, (2020), 185–190. https://doi.org/10.1109/ICMLA51294.2020.00038 |
| [24] |
L. Bai, L. Yao, C. Li, X. Wang, C. Wang, Adaptive graph convolutional recurrent network for traffic forecasting, Adv. Neural Inf. Process. Syst., 33 (2020), 17804–17815. https://doi.org/10.48550/arXiv.2007.02842 doi: 10.48550/arXiv.2007.02842
|
| [25] |
W. Weng, J. Fan, H. Wu, Y. Hu, H. Tian, F. Zhu, et al., A decomposition dynamic graph convolutional recurrent network for traffic forecasting, Pattern Recognit., 142 (2023), 109670. https://doi.org/10.1016/j.patcog.2023.109670. doi: 10.1016/j.patcog.2023.109670
|
| [26] | S. Guo, Y. Lin, N. Feng, C. Song, H. Wan, Attention based spatial-temporal graph convolutional networks for traffic flow forecasting, in Proceedings of the AAAI Conference on Artificial Intelligence, 33 (2019), 922–929. https://doi.org/10.1609/aaai.v33i01.3301922 |
| [27] | Z. Fang, Q. Long, G. Song, K. Xie, Spatial-temporal graph ode networks for traffic flow forecasting, in KDD '21: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, (2021), 364–373. https://doi.org/10.1145/3447548.3467430 |
| [28] | M. Li, Z. Zhu, Spatial-temporal fusion graph neural networks for traffic flow forecasting, in Proceedings of the AAAI Conference on Artificial Intelligence, 35 (2021), 4189–4196. https://doi.org/10.1609/aaai.v35i5.16542 |
| [29] | X. Cai, Y. Zhu, X. Wang, Y. Yao, MambaTS: improved selective state space models for long-term time series forecasting, preprint, arXiv: 2405.16440. https://doi.org/10.48550/arXiv.2405.16440 |
| [30] | D. Liang, H. Zhang, D. Yuan, B. Zhang, M. Zhang, Minusformer: improving time series forecasting by progressively learning residuals, preprint, arXiv: 2402.02332. https://doi.org/10.48550/arXiv.2402.02332 |
| [31] |
W. Lin, Z. Zhang, G. Ren, Y. Zhao, J. Ma, Q. Cao, MGCN: mamba-integrated spatiotemporal graph convolutional network for long-term traffic forecasting, Knowl.-Based Syst., 309 (2025), 112875. https://doi.org/10.1016/j.knosys.2024.112875 doi: 10.1016/j.knosys.2024.112875
|
| [32] |
M. Wu, W. Weng, X. Wang, D. Seng, TSHDNet: temporal-spatial heterogeneity decoupling network for multi-mode traffic flow prediction, Appl. Intell., 55 (2025), 1–14. https://doi.org/10.1007/s10489-024-06218-y doi: 10.1007/s10489-024-06218-y
|
| [33] | B. A. Dai, N. Lyu, Y. Miao, FasterSTS: a faster spatio-temporal synchronous graph convolutional networks for traffic flow forecasting, preprint, arXiv: 2501.00756. https://doi.org/10.48550/arXiv.2501.00756 |
| [34] | T. N. Kipf, M. Welling, Semi-supervised classification with graph convolutional networks, preprint, arXiv: 1609.02907. https://doi.org/10.48550/arXiv.1609.02907 |
Figures(9) / Tables(5)
Ruxin Xue, Jinggui Huang, Zaitang Huang, Bingyan Li. Reconstructed graph spatio-temporal stochastic controlled differential equation for traffic flow forecasting[J]. Electronic Research Archive, 2025, 33(4): 2543-2566. doi: 10.3934/era.2025113
DownLoad: