Abstract
Traffic prediction is an important link in building a smart city, accurate prediction of future traffic conditions can provide a reference for people to travel and facilitate the work of relevant departments, and is vital for urban applications. Nevertheless, due to the cyclical nature of traffic conditions and uncertain factors, as well as the spatial heterogeneity in the traffic network, it is difficult to model the dynamic temporal and spatial correlation, so traffic prediction is facing severe challenges. Existing methods usually use a dynamic graph to model and analyze traffic conditions at a certain moment, lacking consideration of spatial heterogeneity and trend changes in traffic flow, so prediction results are often unsatisfactory. To address the above challenges, we propose a Multi-scale Spatial-temporal Fusion Graph Network for Traffic Prediction framework (MFSTGN). Specifically, we design a spatial-temporal graph convolution network module (STGCN) that dynamically models spatial-temporal correlations while preserving the inherent structure of the traffic network, and describes the trend variation of traffic flows through a kind of trend graph convolution, while modeling the spatial heterogeneity of the traffic network using spatial-temporal embedding. STGCN allows MFSTGN to enjoy a perceptual field that facilitates long-term prediction. In addition, we have developed a Gated Attention mechanism that adaptively fuses cyclical dependence and trend dependence, enabling MFSTGN to enjoy multiple scale information. Experimental results on four datasets for two types of traffic prediction tasks show that MFSTGN outperforms the state-of-the-art baseline.
Similar content being viewed by others
Data availability and access
The PEMS-BAY dataset analysed during the current study are available in the DCRNN repository, https://github.com/liyaguang/DCRNN. The NE-BJ dataset analysed during the current study are available in the DCRNN repository, https://github.com/tsinghua-fib-lab/Traffic-Benchmark. The PEMS04 and PEMS08 datasets analysed during the current study are available in the ASTGNN repository, https://github.com/guoshnBJTU/ASTGNN.
References
Li F, Feng J, Yan H et al (2023) Dynamic graph convolutional recurrent network for traffic prediction: Benchmark and solution. ACM Trans Knowl Discov Data 17(1):1–21
Wang X, Ma Y, Wang Y et al (2020) Traffic flow prediction via spatial temporal graph neural network. Proc Web Conf 2020:1082–1092
Zheng C, Fan X, Wang C et al (2020) GMAN: A graph multi-attention network for traffic prediction. Proc AAAI Conf Artif Intell 34(01):1234–1241
Li Y, Yu R, Shahabi C, et al. (2017) Diffusion convolutional recurrent neural network: Data-driven traffic forecasting. International Conference on Learning Representations
Cascetta E. (2013) Transportation systems engineering: theory and methods. Springer Science Business Media
Lippi M, Bertini M, Frasconi P (2013) Short-term traffic flow forecasting: An experimental comparison of time-series analysis and supervised learning. IEEE Trans Intell Transp Syst 14(2):871–882
Smola AJ, Schölkopf B. (2004) A tutorial on support vector regression. Stat Comput 14:199–222
Kumar SV, Vanajakshi L (2015) Short-term traffic flow prediction using seasonal ARIMA model with limited input data. Eur Transp Res Rev 7(3):1–9
Kumar SV (2017) Traffic flow prediction using Kalman filtering technique. Procedia Eng 187:582–587
Connor JT, Martin RD, Atlas LE (1994) Recurrent neural networks and robust time series prediction. IEEE Trans Neural Netw 5(2):240–254
Hochreiter S, Schmidhuber J (1997) Long short-term memory. Neural Comput 9(8):1735–1780
Kipf T N, Welling M. (2017) Semi-supervised classification with graph convolutional networks. International Conference on Learning Representations
Velickovic P, Cucurull G, Casanova A, et al. (2018) Graph attention networks. International Conference on Learning Representations
Zheng Z, Su D (2014) Short-term traffic volume forecasting: A k-nearest neighbor approach enhanced by constrained linearly sewing principle component algorithm. Transp Res C Emerg Technol 43:143–157
Scarselli F, Gori M, Tsoi AC et al (2009) The graph neural network model. IEEE Trans Neural Netw 20(1):61–80
Li M, Zhu Z (2021) Spatial-temporal fusion graph neural networks for traffic flow forecasting. Proceedings of the AAAI conference on artificial intelligence 35(5):4189–4196
Wu Z, Pan S, Long G, et al. (2019) Graph WaveNet for deep spatial-temporal graph modeling. International Joint Conference on Artificial Intelligence 1907–1913
Bai L, Yao L, Li C et al (2020) Adaptive graph convolutional recurrent network for traffic forecasting. Adv Neural Inf Process Syst 33:17804–17815
Zhang J, Zheng Y, Qi D, et al. (2016) DNN-based prediction model for spatio-temporal data. Proceedings of the 24th ACM SIGSPATIAL international conference on advances in geographic information systems, pp 1–4
Zhang J, Zheng Y, Qi D (2017) Deep spatio-temporal residual networks for citywide crowd flows prediction. Proceedings of the AAAI conference on artificial intelligence 31(1):1655–1661
Yao H, Tang X, Wei H et al (2019) Revisiting spatial-temporal similarity: A deep learning framework for traffic prediction. Proceedings of the AAAI conference on artificial intelligence 33(01):5668–5675
Yao H, Wu F, Ke J, et al. (2018) Deep multi-view spatial-temporal network for taxi demand prediction. Proceedings of the AAAI conference on artificial intelligence, vol 32(1)
Ye J, Zhao J, Ye K et al (2020) How to build a graph-based deep learning architecture in traffic domain: A survey. IEEE Trans Intell Transp Syst 23(5):3904–3924
Chen W, Chen L, Xie Y et al (2020) Multi-range attentive bicomponent graph convolutional network for traffic forecasting. Proceedings of the AAAI conference on artificial intelligence 34(04):3529–3536
Wu Z, Pan S, Long G, et al. (2020) Connecting the dots: Multivariate time series forecasting with graph neural networks. Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pp 753–763
Zhang Q, Chang J, Meng G et al (2020) Spatio-temporal graph structure learning for traffic forecasting. Proceedings of the AAAI conference on artificial intelligence 34(01):1177–1185
Yu B, Yin H, Zhu Z (2018) Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting. International Joint Conference on Artificial Intelligence 3634–3640
Guo S, Lin Y, Feng N et al (2019) Attention based spatial-temporal graph convolutional networks for traffic flow forecasting. Proceedings of the AAAI conference on artificial intelligence 33(01):922–929
Song C, Lin Y, Guo S et al (2020) Spatial-temporal synchronous graph convolutional networks: A new framework for spatial-temporal network data forecasting. Proceedings of the AAAI conference on artificial intelligence 34(01):914–921
Pan Z, Liang Y, Wang W, et al. (2019) Urban traffic prediction from spatio-temporal data using deep meta learning. Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, pp 1720–1730
Song Y, He F, Duan Y et al (2022) A kernel correlation-based approach to adaptively acquire local features for learning 3D point clouds. Comput Aided Des 146:103196
Zhang S, He F (2020) DRCDN: learning deep residual convolutional dehazing networks. Vis Comput 36(9):1797–1808
Zhang J, He F, Duan Y et al (2023) AIDEDNet: Anti-interference and detail enhancement dehazing network for real-world scenes. Front Comput Sci 17(2):172703
Vaswani A, Shazeer N, Parmar N et al (2017) Attention is all you need. Adv Neural Inf Process Syst 30:5998–6008
Zhang Z, Li M, Lin X et al (2019) Multistep speed prediction on traffic networks: A graph convolutional sequence-to-sequence learning approach with attention mechanism. Transp Res C Emerg Technol 105:297–322
Qin Y, Chen H, Jiang G. (2017) A dual-stage attention-based recurrent neural network for time series prediction. International Joint Conference on Artificial Intelligence, pp 2627–2633
Liang Y, Ke S, Zhang J et al (2018) Geoman: Multi-level attention networks for geo-sensory time series prediction. International Joint Conference on Artificial Intelligence 2018:3428–3434
Zhou H, Zhang S, Peng J et al (2021) Informer: Beyond efficient transformer for long sequence time-series forecasting. Proceedings of the AAAI conference on artificial intelligence 35(12):11106–11115
Wu H, Xu J, Wang J et al (2021) Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting. Adv Neural Inf Process Syst 34:22419–22430
Xu D, Dai H, Wang Y, et al. (2019) Road traffic state prediction based on a graph embedding recurrent neural network under the SCATS. Chaos: Interdisc J Nonlinear Sci 29(10):103125
Guo S, Lin Y, Wan H et al (2021) Learning dynamics and heterogeneity of spatial-temporal graph data for traffic forecasting. IEEE Trans Knowl Data Eng 34(11):5415–5428
Lan S, Ma Y, Huang W, et al. (2022) DSTAGNN: Dynamic spatial-temporal aware graph neural network for traffic flow forecasting. International Conference on Machine Learning. PMLR, pp 11906–11917
Wang C, Tian R, Hu J et al (2023) A trend graph attention network for traffic prediction. Inf Sci 623:275–292
Zivot E, Wang J. (2006) Vector autoregressive models for multivariate time series. Model Financ Time Ser S-PLUS® 385–429
Lu H, Huang D, Song Y et al (2020) St-TrafficNet: A spatial-temporal deep learning network for traffic forecasting. Electronics 9(9):1474
Funding
This work is supported in part by the National Natural Science Foundation of China (71961028, 62261050), the Key Research and Development Program of Gansu(22YF7GA171), the Scientific Research Project of the Lanzhou Science and Technology Program (2018-01-58, 2017-4-101), and the Natural Science Foundations of Gansu (21JR7RA119, 21JR7RA208).
Author information
Authors and Affiliations
Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Ran Tian, Chu Wang, Jia Hu and Zhongyu Ma. The first draft of the manuscript was written by Chu Wang and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Ran Tian: Methodology, Conceptualization, Supervision, Project administration, Writing- Reviewing and Editing; Chu Wang: Methodology, Data curation, Software, Writing- Original draft. Jia Hu: Visualization, Investigation, Validation. Zhongyu Ma: Investigation, Writing- Reviewing and Editing.
Corresponding author
Ethics declarations
Ethical and informed consent for data used
All authors strictly understand and comply with the ethical and informed consent for data used
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Tian, R., Wang, C., Hu, J. et al. MFSTGN: a multi-scale spatial-temporal fusion graph network for traffic prediction. Appl Intell 53, 22582–22601 (2023). https://doi.org/10.1007/s10489-023-04703-4
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-023-04703-4