Academic Journal of Computing & Information Science, 2024, 7(5); doi: 10.25236/AJCIS.2024.070511.
Zhaorui Ge1, Shuai Dai2
1People's Public Security University of China, Beijing, 100038, China
2Road Traffic Safety Research Center of the Ministry of Public Security, Beijing, 100062, China
Traffic flow prediction plays an important role in smart city construction. Aiming at the problem that it is difficult to extract the spatio-temporal dynamic correlation of traffic flow, a traffic flow prediction model based on wavelet analysis, 2D convolutional neural network and graph convolutional neural network (WAC-GCN) was proposed. Firstly, the spatial correlation heat map was constructed by nodes to find the spatial correlation characteristics between different nodes. Secondly, the control variable method is used to adjust the parameters of the model, and the Early Stopping technique is introduced to improve the generalization performance of the model and reduce the waste of time and resources. Finally, the test set was used and different modules were eliminated to obtain the prediction results of the corresponding traffic flow prediction model. The experimental results on the real highway data set show that the proposed network model has better accuracy than the baseline model with some modules removed.
Traffic flow prediction; Wavelet transform; Graph convolutional neural network; 2D convolutional neural network; Spatial characteristics
Zhaorui Ge, Shuai Dai. A Spatio-temporal Traffic Flow Forecasting Method Based on WAC-GCN. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 5: 87-93. https://doi.org/10.25236/AJCIS.2024.070511.
[1] Helfert, Markus, et al., eds. Smart Cities, Green Technologies, and Intelligent Transport Systems: 4th International Conference, SMARTGREENS 2015, and 1st International Conference VEHITS 2015, Lisbon, Portugal, May 20-22, 2015, Revised Selected Papers. Vol. 579. Springer, 2016.
[2] Ma, Dai, et al. "Learning traffic as images: A deep convolutional neural network for large-scale transportation network speed prediction." Sensors 17.4 (2017): 818.
[3] Zhang, Junbo, et al. "Flow prediction in spatio-temporal networks based on multitask deep learning." IEEE Transactions on Knowledge and Data Engineering 32.3 (2019): 468-478.
[4] Li, Chensheng, et al. "Scalable graph convolutional networks with fast localized spectral filter for directed graphs." IEEE Access 8 (2020): 105634-105644.
[5] Zhang, Xiyue, et al. "Spatial-Temporal Convolutional Graph Attention Networks for Citywide Traffic Flow Forecasting." ACM (2020).
[6] Zhao L, Song Y, Zhang C, et al. T-gcn: A temporal graph convolutional network for traffic prediction [J]. IEEE transactions on intelligent transportation systems, 2019, 21(9): 3848-3858.
[7] Sun L, Liu M, Liu G, et al. FD-TGCN: Fast and dynamic temporal graph convolution network for traffic flow prediction [J]. Information Fusion, 2024: 102291.
[8] Li, Zhishuai, et al. "A hybrid deep learning approach with GCN and LSTM for traffic flow prediction." 2019 IEEE intelligent transportation systems conference (ITSC). IEEE, 2019.
[9] Hussain, Basharat, et al. "Intelligent traffic flow prediction using optimized GRU model." IEEE Access 9 (2021): 100736-100746.
[10] Kushchenko, L. E., S. V. Kushchenko, and A. N. Novikov. "The Application of Wavelet Analysis to Study the Characteristics of the Traffic Flow." 2020 International Multi-Conference on Industrial Engineering and Modern Technologies (FarEastCon). IEEE, 2020.
[11] Zhou, Yan, et al. "Encrypted network traffic identification based on 2d-cnn model." 2021 22nd Asia-Pacific Network Operations and Management Symposium (APNOMS). IEEE, 2021.