Abstract
The traffic flow prediction of expressway is highly nonlinear and complex, so it is technically and operationally difficult to predict it accurately. In order to predict the traffic flow of expressway and alleviate its traffic congestion, a multi-component spatial–temporal graph convolution network model based on attention mechanism (MSTA-GCN) is proposed, which combines GCN, standard CNN and attention mechanism. Firstly, near period, daily period and weekly period are modeled to capture the nonlinear correlation characteristics in multivariable urban spatial–temporal series. Then input the sample data into the integrated model for training and extract the characteristics of traffic flow data. Finally, the model is tested on California Expressway PeMS D7 dataset. The experimental results show that MSTA-GCN is superior to classical shallow learning model, benchmark deep learning model and other traffic flow deep learning models, and has lower prediction error.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data availability statement
Data will be made available on reasonable request.
References
Asif MT, Mitrovic N, Dauwels J et al (2016) Matrix and tensor based methods for missing data estimation in large traffic networks. IEEE Trans Intell Transp Syst 17(7):1816–1825
Bae B, Kim H, Lim H et al (2018) Missing data imputation for traffic flow speed using spatio-temporal cokriging. Transp Res Part C Emerg Technol 88:124–139
Chiang WL, Liu XQ, Si S, et al (2019) Cluster-GCN: An efficient algorithm for training deep and large graph convolutional networks. In: Proc of the 25th ACM SIGKDD Int Conf on knowledge discovery & data mining 19(8):257–266
Du SD, Li TR, Yang Y et al (2020) A sequnce-to sequence spatial-temporal attention learing model for urban traffic flow prediction. J Comput Res Dev 57(8):1715–1728
Feng N, Guo SN, Song C et al (2019) Multi-component spatial-temporal graph convolution networks for traffic flow forecasting. J Softw 30(3):759–769
Gui Z, Li Z, Guo L (2020) Short-term traffic flow prediction based on ACGRU model. Comput Eng Appl 56(21):260–265
Karafyllis I, Theodosis D, Papageorigiou M (2022) Analysis and control of a non-local PDE traffic flow model. Int J Control 95(3):660–678
Lee N, Choi W, Vernaza P, et al (2017) Desire: Distant future prediction in dynamic scenes with interacting agents. In: IEEE. Proceedings of the IEEE Conference on computer vision and pattern recognition. New York: IEEE, pp 336–345
Li LJ (2020) Research on spatial temporal prediction model of traffic flow based on attention mechanism. South China University of Technology, Guanzhou
Li TY, Wang T, Zhang YQ (2021) Highway traffic flow prediction model with multi-features. J Transp Syst Eng Inf Technol 21(3):101–121
Liao L, Hu Z, Zheng Y, et al (2022) An improved dynamic Chebyshev graph convolution network for traffic flow prediction with spatial-temporal attention. Appl Intell 3(19):1–13
Liu JX, Chen SC (2019) Non-stationary multivariate time series prediciton with MIX gated unit. J Comput Res Dev 56(8):1642–1651
Ni QJ, Peng WQ, Zhang ZZ et al (2022) Spatial-temporal graph neural network for traffic flow prediction based on information enhanced transmission. J Comput Res Dev 59(2):282–293
Ning T, An L, Duan XD (2021) Optimization of cold chain distribution path of fresh agriculture products under caorbon tax mechanism: a case study in China. J Intell Fuzzy Syst 40(6):10549–10558
Shang P, Liu X, Yu C et al (2022) A new ensemble deep graph reinforcement learning network for spatio-temporal traffic volume forecasting in a freeway network. Digit Signal Process 123:103419
Tao N, Shi S, Peng Z et al (2021) Distribution management decision model for VRPDP under changes of customer distribution demand. J Ambient Intell Humaniz Comput 2(2):2053–2063
Woodard D, Nogin G, Koch P et al (2017) Predicting travel time reliability using mobile phone GSP data. Transp Res Part C Emerg Technol 75:30–44
Xue YR, Wu R, Liu JF (2021) Crowd evacuation guidance based on combined action-space deep reinforement learning. J Harbin Inst Technol 53(8):29–38
Yan Z, Yu C-C, Han L et al (2019) Short-term traffic flow forecasting method based on CNN+LSTM. Comput Eng Des 40(09):2620–2624
Yan X, Fan X-l, Zheng C-p et al (2020) Urban traffic flow prediction algorithm based on graph convolutional neural networks. J Zhejiang Univ (engineering Science) 54(6):1147–1155
Yang D, Li S, Peng Z et al (2019) MF-CNN: traffic flow prediction using convolutional neural network and multi-features fusion. IEICE Trans Inf Syst 102(8):1526–1536
Zhang JJ, Wang YB, Long MS et al (2020) Predictive recurrent network for seasonal spatial-temporal data and its application in urban computing. Chin J Comput 43(2):286–302
Zhang W-b, Zhang P-l, Su Z-y et al (2021) Missing data repairs for road network traffic flow with self-attention graph auto-encoder networks. J Transp Syst Eng Inf Technol 21(4):90–98
Zhou J, Gui GQ, Hu SD et al (2020) Graph neural networks: a review of methods and applications. AI Open 1:57–81
Acknowledgements
This work is partially supported by Research Project on Economic and Social Development of Liaoning Province (2022lsljdybkt-014), Science and technology innovation fund program of Dalian (2021JJ13SN81), Scientific and Research Project of Education Department of Liaoning Province (No. L2020006) and Research project of China Federation of logistics and procurement (Grant No.:2022CSLKT3-021).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
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
Ning, T., Wang, J. & Duan, X. Research on expressway traffic flow prediction model based on MSTA-GCN. J Ambient Intell Human Comput 14, 9317–9328 (2023). https://doi.org/10.1007/s12652-022-04431-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12652-022-04431-6