IEEE/CAA Journal of Automatica Sinica
Citation:  Z. B. Wei, H. X. Zhao, Z. S. Li, X. J. Bu, Y. Y. Chen, X. Q. Zhang, Y. S. Lv, and F.Y. Wang, “STGSA: A novel spatialtemporal graph synchronous aggregation model for traffic prediction,” IEEE/CAA J. Autom. Sinica, vol. 10, no. 1, pp. 226–238, Jan. 2023. doi: 10.1109/JAS.2023.123033 
The success of intelligent transportation systems relies heavily on accurate traffic prediction, in which how to model the underlying spatialtemporal information from traffic data has come under the spotlight. Most existing frameworks typically utilize separate modules for spatial and temporal correlations modeling. However, this stepwise pattern may limit the effectiveness and efficiency in spatialtemporal feature extraction and cause the overlook of important information in some steps. Furthermore, it is lacking sufficient guidance from prior information while modeling based on a given spatial adjacency graph (e.g., deriving from the geodesic distance or approximate connectivity), and may not reflect the actual interaction between nodes. To overcome those limitations, our paper proposes a spatialtemporal graph synchronous aggregation (STGSA) model to extract the localized and longterm spatialtemporal dependencies simultaneously. Specifically, a tailored graph aggregation method in the vertex domain is designed to extract spatial and temporal features in one graph convolution process. In each STGSA block, we devise a directed temporal correlation graph to represent the localized and longterm dependencies between nodes, and the potential temporal dependence is further finetuned by an adaptive weighting operation. Meanwhile, we construct an elaborated spatial adjacency matrix to represent the road sensor graph by considering both physical distance and node similarity in a datadriven manner. Then, inspired by the multihead attention mechanism which can jointly emphasize information from different representation subspaces, we construct a multistream module based on the STGSA blocks to capture global information. It projects the embedding input repeatedly with multiple different channels. Finally, the predicted values are generated by stacking several multistream modules. Extensive experiments are constructed on six realworld datasets, and numerical results show that the proposed STGSA model significantly outperforms the benchmarks.
[1] 
F.Y. Wang, “Parallel control and management for intelligent transportation systems: Concepts, architectures, and applications,” IEEE Trans. Intelligent Transportation Systems, vol. 11, no. 3, pp. 630–638, 2010. doi: 10.1109/TITS.2010.2060218

[2] 
J. P. Zhang, F.Y. Wang, K. F. Wang, W.H. Lin, X. Xu, and C. Chen, “Datadriven intelligent transportation systems: A survey,” IEEE Trans. Intelligent Transportation Systems, vol. 12, no. 4, pp. 1624–1639, 2011. doi: 10.1109/TITS.2011.2158001

[3] 
F.Y. Wang, “Scanning the issue and beyond: Computational transportation and transportation 5.0,” IEEE Trans. Intelligent Transportation Systems, vol. 5, no. 15, pp. 1861–1868, 2014.

[4] 
M. Veres and M. Moussa, “Deep learning for intelligent transportation systems: A survey of emerging trends,” IEEE Trans. Intelligent Transportation Systems, vol. 21, no. 8, pp. 3152–3168, 2019.

[5] 
N. Zhang, F.Y. Wang, F. H. Zhu, D. B. Zhao, and S. M. Tang, “Dynacas: Computational experiments and decision support for its,” IEEE Intelligent Systems, vol. 23, no. 6, pp. 19–23, 2008. doi: 10.1109/MIS.2008.101

[6] 
Y. Lee, H. Jeon, and K. Sohn, “Predicting shortterm traffic speed using a deep neural network to accommodate citywide spatiotemporal correlations,” IEEE Trans. Intelligent Transportation Systems, vol. 22, no. 3, pp. 1435–1448, 2021. doi: 10.1109/TITS.2020.2970754

[7] 
Z. S. Li, G. Xiong, Y. L. Tian, Y. S. Lv, Y. Y. Chen, P Hui, and X. Su, “A multistream feature fusion approach for traffic prediction,” IEEE Trans. Intelligent Transportation Systems, vol. 23, no. 2, pp. 1456–1466, Feb. 2022.

[8] 
C. Song, Y. F. Lin, S. N. Guo, and H. Y. Wan, “Spatialtemporal synchronous graph convolutional networks: A new framework for spatialtemporal network data forecasting,” in Proc. AAAI Conf. Artificial Intelligence, 2020, pp. 914–921.

[9] 
Y. S. Lv, Y. J. Duan, W. W. Kang, Z. X. Li, and F.Y. Wang, “Traffic flow prediction with big data: A deep learning approach,” IEEE Trans. Intelligent Transportation Systems, vol. 16, no. 2, pp. 865–873, 2014.

[10] 
N. G. Polson and V. O. Sokolov, “Deep learning for shortterm traffic flow prediction,” Transportation Research Part C: Emerging Technologies, vol. 79, pp. 1–17, 2017. doi: 10.1016/j.trc.2017.02.024

[11] 
Y. G. Li, R. Yu, C. Shahabi, and Y. Liu, “Diffusion convolutional recurrent neural network: Datadriven traffic forecasting,” in Proc. Int. Conf. Learning Representations, 2018.

[12] 
B. Yu, H. Yin, and Z. Zhu, “Spatiotemporal graph convolutional networks: a deep learning framework for traffic forecasting,” in Proc. 27th Int. Joint Conf. Artificial Intelligence, 2018, pp. 3634–3640.

[13] 
A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention is all you need,” in NIPS, 2017.

[14] 
J. H. Guo, W. Huang, and B. M. Williams, “Adaptive kalman filter approach for stochastic shortterm traffic flow rate prediction and uncertainty quantification,” Transportation Research Part C: Emerging Technologies, vol. 43, pp. 50–64, 2014. doi: 10.1016/j.trc.2014.02.006

[15] 
M. Van Der Voort, M. Dougherty, and S. Watson, “Combining kohonen maps with arima time series models to forecast traffic flow,” Transportation Research Part C: Emerging Technologies, vol. 4, no. 5, pp. 307–318, 1996. doi: 10.1016/S0968090X(97)829038

[16] 
R. M. Li, C. Y. Jiang, F. H. Zhu, and X. L. Chen, “Traffic flow data forecasting based on interval type2 fuzzy sets theory,” IEEE/CAA J. Autom. Sinica, vol. 3, no. 2, pp. 141–148, 2016. doi: 10.1109/JAS.2016.7451101

[17] 
R. M. Li, Y. F. Huang, and J. Wang, “Longterm traffic volume prediction based on kmeans gaussian interval type2 fuzzy sets,” IEEE/CAA J. Autom. Sinica, vol. 6, no. 6, pp. 1344–1351, 2019.

[18] 
L. Li, X. Q. Chen, and L. Zhang, “Multimodel ensemble for freeway traffic state estimations,” IEEE Trans. Intelligent Transportation Systems, vol. 15, no. 3, pp. 1323–1336, 2014. doi: 10.1109/TITS.2014.2299542

[19] 
V. Le Guen and N. Thome, “Probabilistic time series forecasting with shape and temporal diversity,” Advances in Neural Information Processing Systems, vol. 33, 2020.

[20] 
J. Schmidhuber and S. Hochreiter, “Long shortterm memory,” Neural Comput, vol. 9, no. 8, pp. 1735–1780, 1997. doi: 10.1162/neco.1997.9.8.1735

[21] 
R. Dai, S. K. Xu, Q. Gu, C. G. Ji, and K. K. Liu, “Hybrid spatiotemporal graph convolutional network: Improving traffic prediction with navigation data,” in Proc. 26th ACM SIGKDD Int. Conf. Knowledge Discovery & Data Mining, 2020, pp. 3074–3082.

[22] 
Y. N. Dauphin, A. Fan, M. Auli, and D. Grangier, “Language modeling with gated convolutional networks,” in Proc. Int. Conf. Machine Learning. PMLR, 2017, pp. 933–941.

[23] 
Z. N. Yuan, X. Zhou, and T. B. Yang, “HeteroConvLSTM: A deep learning approach to traffic accident prediction on heterogeneous spatiotemporal data,” in Proc. 24th ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, 2018, pp. 984–992.

[24] 
Y. Seo, M. Defferrard, P. Vandergheynst, and X. Bresson, “Structured sequence modeling with graph convolutional recurrent networks,” in Proc. Int. Conf. Neural Information Processing, 2018, pp. 362–373.

[25] 
Y. Y. Chen, H. Y. Chen, P. J. Ye, Y. S. Lv, and F.Y. Wang, “Acting as a decision maker: Trafficconditionaware ensemble learning for traffic flow prediction,” IEEE Trans. Intelligent Transportation Systems, vol. 23, no. 4, pp. 1–11, 2020.

[26] 
Y. L. Lin, X. Y. Dai, L. Li, and F.Y. Wang, “Pattern sensitive prediction of traffic flow based on generative adversarial framework,” IEEE Trans. Intelligent Transportation Systems, vol. 20, no. 6, pp. 2395–2400, 2018.

[27] 
S. Y. Hao, D.H. Lee, and D. Zhao, “Sequence to sequence learning with attention mechanism for shortterm passenger flow prediction in largescale metro system,” Transportation Research Part C: Emerging Technologies, vol. 107, pp. 287–300, 2019. doi: 10.1016/j.trc.2019.08.005

[28] 
L. Cai, K. Janowicz, G. C. Mai, B. Yan, and R. Zhu, “Traffic transformer: Capturing the continuity and periodicity of time series for traffic forecasting,” Transactions in GIS, vol. 24, no. 3, pp. 736–755, 2020. doi: 10.1111/tgis.12644

[29] 
Z. Y. Pan, Y. X. Liang, W. F. Wang, Y. Yu, Y. Zheng, and J. B. Zhang, “Urban traffic prediction from spatiotemporal data using deep meta learning,” in Proc. 25th ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, 2019, pp. 1720–1730.

[30] 
J. B. Estrach, W. Zaremba, A. Szlam, and Y. LeCun, “Spectral networks and deep locally connected networks on graphs,” in Proc. 2nd Int. Conf. Learning Representations, ICLR, 2014.

[31] 
M. Defferrard, X. Bresson, and P. Vandergheynst, “Convolutional neural networks on graphs with fast localized spectral filtering,” in NIPS, 2016.

[32] 
T. N. Kipf and M. Welling, “Semisupervised classification with graph convolutional networks,” arXiv preprint arXiv: 1609.02907, 2016.

[33] 
W. L. Hamilton, R. Ying, and J. Leskovec, “Inductive representation learning on large graphs,” in Proc. 31st Int. Conf. Neural Information Processing Systems, 2017, pp. 1025–1035.

[34] 
J. Atwood and D. Towsley, “Diffusionconvolutional neural networks,” in Advances in Neural Information Processing Systems, 2016, pp. 1993–2001.

[35] 
P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y. Bengio, “Graph attention networks,” in Proc. 7th Int. Conf. Learning Representations, 2018.

[36] 
X. Y. Wang, Y. Ma, Y. Q. Wang, W. Jin, X. Wang, J. L. Tang, C. Y. Jia, and J. Yu, “Traffic flow prediction via spatial temporal graph neural network,” in Proc. Web Conf., 2020, pp. 1082–1092.

[37] 
G. Q. Liu, Y. L. Wu, D. Zhao, and H. C. Zhou, “Timeadaptive graph convolutional network for traffic prediction,” in Proc. 5th Int. Conf. Deep Learning Technologies, 2021, pp. 81–86.

[38] 
W. Zhang, F. H. Zhu, Y. S. Lv, C Tan, W. Liu, X. Zhang, and F.Y. Wang “AdapGL: An adaptive graph learning algorithm for traffic prediction based on spatiotemporal neural networks,” Transportation Research Part C: Emerging Technologies, vol. 139, p. 103659, 2022.

[39] 
A. T. Khan, X. W. Cao, Z. Li, and S. Li, “Enhanced beetle antennae search with zeroing neural network for online solution of constrained optimization,” Neurocomputing, vol. 447, pp. 294–306, 2021. doi: 10.1016/j.neucom.2021.03.027

[40] 
A. T. Khan, S. Li, and X. W. Cao, “Human guided cooperative robotic agents in smart home using beetle antennae search,” Science China Information Sciences, vol. 65, no. 2, pp. 1–17, 2022.

[41] 
A. T. Khan, S. Li, and Z. Li, “Obstacle avoidance and modelfree tracking control for home automation using bioinspired approach,” Advanced Control for Applications: Engineering and Industrial Systems, vol. 4, no. 1, p. e63, 2022.

[42] 
Z. C. Zhang, X. Lin, M. Li, and Y. H. Wang, “A customized deep learning approach to integrate networkscale online traffic data imputation and prediction,” Transportation Research Part C: Emerging Technologies, vol. 132, p. 103372, 2021. doi: 10.1016/j.trc.2021.103372

[43] 
D. Wu and X. Luo, “Robust latent factor analysis for precise representation of highdimensional and sparse data,” IEEE/CAA J. Autom. Sinica, vol. 8, no. 4, pp. 796–805, 2020.

[44] 
H. Wu, X. Luo, and M. C. Zhou, “Advancing nonnegative latent factorization of tensors with diversified regularizations,” IEEE Trans. Services Computing, vol. 15, no. 3, pp. 1334–1344, 2020.

[45] 
X. Luo, Y. Yuan, S. L. Chen, N. Y. Zeng, and Z. D. Wang, “Positiontransitional particle swarm optimizationincorporated latent factor analysis,” IEEE Trans. Knowledge and Data Engineering, vol. 34, no. 8, pp. 3958–3970, 2022.

[46] 
X. Geng, Y. G. Li, L. Y. Wang, L. Y. Zhang, Q. Yang, J. P. Ye, and Y. Liu, “Spatiotemporal multigraph convolution network for ridehailing demand forecasting,” in Proc. AAAI Conf. Artificial Intelligence, 2019, pp. 3656–3663.

[47] 
R. Y. Li, S. Wang, F. Y. Zhu, and J. Z. Huang, “Adaptive graph convolutional neural networks,” in Proc. AAAI Conf. Artificial Intelligence, 2018.

[48] 
Z. H. Wu, S. R. Pan, G. D. Long, J. Jiang, and C. Q. Zhang, “Graph WaveNet for deep spatialtemporal graph modeling.” in Proc. 28th Int. Joint Conf. Artificial Intelligence, 2019, pp. 1907–1913.

[49] 
C. P. Zheng, X. L. Fan, C. Wang, and J. Z. Qi, “GMAN: A graph multiattention network for traffic prediction,” in Proc. AAAI Conf. Artificial Intelligence, 2020, pp. 1234–1241.

[50] 
L. Bai, L. N. Yao, C. Li, X. Z. Wang, and C. Wang, “Adaptive graph convolutional recurrent network for traffic forecasting,” Advances in Neural Information Processing Systems, vol. 33, 2020.

[51] 
M. Z. Li and Z. X. Zhu “Spatialtemporal fusion graph neural networks for traffic flow forecasting,” arXiv preprint arXiv: 2012.09641, 2020.

[52] 
I. Sutskever, O. Vinyals, and Q. V. Le, “Sequence to sequence learning with neural networks,” Advances in Neural Information Processing Systems, vol. 27, pp. 3104–3112, 2014.
