|
|
基于优化图注意力网络的解耦式交通流预测仿真模型
|
Abstract:
高效的交通流预测对提升智能交通系统的性能至关重要。其关键挑战在于精准地建模复杂的交通动态，并全面地捕获交通流数据的时空相关性。大多数时空网络无法单独处理交通流时间序列中非平稳部分的分布变化，并缺少在空间相关性方面建模的高效算法。为此，本文设计一种新的基于优化图注意力网络的解耦式交通流预测仿真模型(D-EFGAT)。该模型设计了一种新的解耦–融合框架，利用二次分解将复杂的交通流数据解耦为稳定趋势和波动事件序列，通过双尺度时空编码网络分别对趋势和事件进行建模，最后进行自适应融合。此外，在图注意力网络中引入了注意力筛选机制和动态时空图编码，更加高效地建模动态空间相关性。利用美国PeMS的交通流数据集进行仿真实验，仿真结果表明D-EFGAT与基线模型相比具有最优的预测性能。
Efficient traffic flow prediction is crucial to improve the performance of intelligent transport systems. The main challenge is to accurately model complex traffic dynamics and comprehensively capture the spatio-temporal correlation of traffic flow data. Many spatio-temporal networks struggle to handle the distributional variations of the non-stationary part of the traffic flow time series and lack efficient algorithms for modeling spatial correlation. Therefore, a new decoupled traffic flow prediction simulation model (D-EFGAT) based on optimized graph attention networks was proposed. The model utilized a new decoupling-fusion framework that employed quadratic decomposition to separate complex traffic flow data into stable trends and fluctuating event sequences. The trends and events were then modeled separately using a dual-scale spatio-temporal coding network, and finally, adaptive fusion was performed. Furthermore, an attention screening mechanism and dynamic spatio-temporal graph encoding were introduced into the graph attention network to model dynamic spatial correlations more efficiently. Simulation experiments are conducted using the traffic flow dataset from PeMS in the U.S. The simulation results show that D-EFGAT has the optimal prediction performance compared to the baseline model.
| [1] | Fang, Y., Qin, Y., Luo, H., et al. (2023) When Spatio-Temporal Meet Wavelets: Disentangled Traffic Forecasting via Efficient Spectral Graph Attention Networks. 2023 IEEE 39th International Conference on Data Engineering (ICDE), Anaheim, 3-7 April 2023, 517-529. https://doi.org/10.1109/ICDE55515.2023.00046 |
| [2] | Li, Y., Yu, R., Shahabi, C., et al. (2017) Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting. arXiv preprint arXiv:1707.01926. https://doi.org/10.48550/arXiv.1707.01926 |
| [3] | Wu, Z., Pan, S., Long, G., et al. (2019) Graph Wavenet for Deep Spatial-Temporal Graph Modeling. Proceedings of the 28th International Joint Conference on Artificial Intelligence, Macao, 10-16 August 2019, 1907-1913. https://doi.org/10.24963/ijcai.2019/264 |
| [4] | Bai, L., Yao, L., Li, C., et al. (2020) Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting. Advances in Neural Information Processing Systems, 33, 17804-17815. |
| [5] | Li, M. and Zhu, Z. (2020) Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting. Proceedings of the AAAI Conference on Artificial Intelligence, 35, 4189-4196. https://doi.org/10.1609/aaai.v35i5.16542 |
| [6] | 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, 922-929. https://doi.org/10.1609/aaai.v33i01.3301922 |
| [7] | Huang, R., Huang, C., Liu, Y., et al. (2020) LSGCN: Long Short-Term Traffic Prediction with Graph Convolutional Networks. Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, Yokohama, 7-15 January 2021, 2355-2361. https://doi.org/10.24963/ijcai.2020/326 |
| [8] | Feng, A. and Tassiulas, L. (2022) Adaptive Graph Spatial-Temporal Transformer Network for Traffic Forecasting. Proceedings of the 31st ACM International Conference on Information & Knowledge Management, Atlanta GA, 17-21 October 2022, 3933-3937. https://doi.org/10.1145/3511808.3557540 |
| [9] | Bengio, Y., Courville, A. and Vincent, P. (2013) Representation Learning: A Review and New Perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35, 1798-1828. https://doi.org/10.1109/TPAMI.2013.50 |
| [10] | Fu, W., Zhang, K., Wang, K., et al. (2021) A Hybrid Approach for Multi-Step Wind Speed Forecasting Based on Two-Layer Decomposition, Improved Hybrid DE-HHO Optimization and KELM. Renewable Energy, 164, 211-229. https://doi.org/10.1016/j.renene.2020.09.078 |
| [11] | Pan, H., Yang, Y., Li, X., et al. (2019) Symplectic Geometry Mode Decomposition and Its Application to Rotating Machinery Compound Fault Diagnosis. Mechanical Systems and Signal Processing, 114, 189-211. https://doi.org/10.1016/j.ymssp.2018.05.019 |
| [12] | Vaswani, A., Shazeer, N., Parmar, N., et al. (2017) Attention Is All You Need. Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, 4-9 December 2017, 6000-6010. |
| [13] | Zivot, E. and Wang, J. (2006) Vector Autoregressive Models for Multivariate Time Series. In: Zivot, E. and Wang, J., Eds., Modeling Financial Time Series with S-PLUS？, Springer, New York, 369-413. |
| [14] | Rucker, H., Burges, C.J., Kaufman, L., et al. (1996) Support Vector Regression Machines. Proceedings of the 9th International Conference on Neural Information Processing Systems, Denver, 3-5 December 1996, 155-161. |
| [15] | Hochreiter, S. and Schmidhuber J. (1997) Long Short-Term Memory. Neural Computation, 9, 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735 |
| [16] | Li, C.L., Cui, Z., Zheng, W.M., et al. (2018) Spatio-Temporal Graph Convolution for Skeleton Based Action Recognition. Proceedings of the AAAI Conference on Artificial Intelligence, 32. https://doi.org/10.1609/aaai.v32i1.11776 |
| [17] | Song, C., Lin, Y.F., Guo, S.N., 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, 914-921. https://doi.org/10.1609/aaai.v34i01.5438 |
