基于图神经网络的交通速度迁移预测

交通流预测是智慧城市建设的一个重要方向，可以为交通调度提供重要依据。 随着数据和技术的发展，交通领域学者提出了智能交通系统，发挥大数据对交通 场景的调度作用。然而，由于地域和资金等差异，不同城市的发展存在不平衡的情 况。现有的研究很少关注到欠发达地区的发展窘境。为了解决这个问题，本文充 分利用已有的城市交通数据流信息，结合深度学习方法，提出了相应的解决方案。
本文提出了两种新的交通预测模型，分别是基于图网络的图划分速度预测模 型(TEEPEE)和基于互信息时间聚类的交通预测模型(TrafficTL)。
TEEPEE 模型使用图划分机制将源城市与目标城市的子区域进行匹配，从而 拉近区域距离。接着，利用图神经卷积网络和门控时序卷积网络提取时间-空间信 息，以此为目标城市各子区域提供数据进行训练学习。这种方法具有较强的区域 关联性，能够很好地利用区域之间的相似性。同时，TEEPEE 还能够通过节点关系 的优化，将预测性能优化了 1.3% 左右。
相比之下，TrafficTL 模型则使用互信息对时间-速度趋势进行相似性聚类，从 而更好地分类和分区子区域的数据。此外，TrafficTL 还采用了图重构模块对节点 关系进行重新构建，以减少先验知识和数据收集中出现的错误。通过这些手段， TrafficTL 不仅具有较高的预测准确性，而且能够更好地适应新的数据集。在实验 中，TrafficTL 还采用了集成机制来减少不同区域间存在的共享参数优化方向冲突 问题，并在整个模型中减少负迁移的出现。通过这些方法，TrafficTL 可以更好地 预测城市交通状况，较之现存的模型，性能改善了 8% 到 25%。
本文提出了两种新的交通预测模型，分别是 TEEPEE 和 TrafficTL。这些模型 采用了不同的方法来利用时间-空间信息和节点关系，从而提高预测准确性和适应性。本文的实验结果表明，这些模型在预测城市交通状况方面具有较高的准确性 和实用性，可以为城市规划和交通管理提供有价值的决策支持。为欠发达地区未 来发展方向提供了潜在的方法支撑。

[1] 中共中央国务院. 交通强国建设纲要[EB/OL]. 2019. http://xxgk.mot.gov.cn/2020/jigou/zcy js/202006/t20200623_3307512.html.
[2] 赵楠, 真虹. 基于第三次工业革命的货物运输结构演变趋势[J]. 中国航海, 2020(116- 120+127).
[3] 伍蕾, 谢波. “技术”与“人本”理念下未来城市的空间发展模式[J]. 规划师, 2020(14- 19+44).
[4] 宗德新, 杨丹. 第四次工业革命影响下的城市发展模式探究[J/OL]. 中国名城, 2022(28-34). DOI: 10.19924/j.cnki.1674-4144.2022.07.004.
[5] 王超深, 谭敏, 冯田. 欠发达地区城市交通规划编制历程、演进特征:回顾及反思——以 四川省为例[C]//绿色·智慧·融合——2021/2022 年中国城市交通规划年会论文集. 2022: 35-44.
[6] 杨利柯, 张文静. 提升许昌新型城镇化城市交通水平的研究[J]. 产业与科技论坛, 2022 (54-56).
[7] 郑宇. 城市计算概述[J/OL]. 武汉大学学报 (信息科学版), 2015(1-13). DOI: 10.13203/j.whu gis20140718.
[8] KUMAR N, RAUBAL M. Applications of deep learning in congestion detection, prediction and alleviation: A survey[J]. Transportation Research Part C: Emerging Technologies, 2021, 133: 103432.
[9] LITMAN T. Smart congestion relief: Comprehensive analysis of traffic congestion costs and congestion reduction benefits[Z]. 2016.
[10] CHEN C, LI K, TEO S G, et al. Citywide traffic flow prediction based on multiple gated spatio- temporal convolutional neural networks[J]. ACM Transactions on Knowledge Discovery from Data (TKDD), 2020, 14(4): 1-23.
[11] GUO S, LIN Y, WAN H, et al. Learning Dynamics and Heterogeneity of Spatial-Temporal Graph Data for Traffic Forecasting[J]. IEEE Transactions on Knowledge and Data Engineering, 2021: 1-1.
[12] LI J, GUO F, WANG Y, et al. Short-term Traffic Prediction with Deep Neural Networks and Adaptive Transfer Learning[J]. 2020 IEEE 23rd International Conference on Intelligent Trans- portation Systems (ITSC), 2020: 1-6.
[13] YIN X, WU G, WEI J, et al. A Comprehensive Survey on Traffic Prediction: abs/2004.08555 [A]. 2020.
[14] HUANG W, SONG G, HONG H, et al. Deep Architecture for Traffic Flow Prediction: Deep Belief Networks With Multitask Learning[J]. IEEE Transactions on Intelligent Transportation Systems, 2014, 15: 2191-2201.
[15] LV Y, DUAN Y, KANG W, et al. Traffic Flow Prediction With Big Data: A Deep Learning Approach[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16: 865-873.
[16] CHAO K H, CHEN P Y. An Intelligent Traffic Flow Control System Based on Radio Frequency Identification and Wireless Sensor Networks[J]. International Journal of Distributed Sensor Networks, 2014, 10.
[17] XIA Y, SHI X, SONG G, et al. Towards improving quality of video-based vehicle counting method for traffic flow estimation[J]. Signal Process., 2016, 120: 672-681.
[18] WANG P, LAI J, HUANG Z, et al. Estimating Traffic Flow in Large Road Networks Based on Multi-Source Traffic Data[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22: 5672-5683.
[19] AHMED M S, COOK A R. Analysis of Freeway Traffic Time-Series Data by Using Box-Jenkins Techniques[J]. Transportation Research Record, 1979.
[20] 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, 2020, 21: 3848-3858.
[21] YU B, YIN H, ZHU Z. Spatio-temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting[C]//Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI). 2018.
[22] LI Y, YU R, SHAHABI C, et al. Diffusion Convolutional Recurrent Neural Network: Data- Driven Traffic Forecasting[A]. 2018.
[23] PAN S J, YANG Q. A Survey on Transfer Learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22: 1345-1359.
[24] WEI Y, ZHENG Y, YANG Q. Transfer knowledge between cities[C]//Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining. San Francisco, CA, USA, 2016: 1905-1914.
[25] WANG L, GENG X, MA X, et al. Ridesharing car detection by transfer learning[J]. Artif. Intell., 2019, 273: 1-18.
[26] GUO B, LI J, ZHENG V W, et al. CityTransfer: Transferring Inter- and Intra-City Knowledge for Chain Store Site Recommendation Based on Multi-Source Urban Data[J]. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., 2018, 1(4): 1-23.
[27] XU F F, LIN B Y, LU Q, et al. Cross-region traffic prediction for China on OpenStreetMap[C]// Proceedings of the 9th ACM SIGSPATIAL International Workshop on Computational Trans- portation Science. 2016: 37-42.
[28] LIN B Y, XU F F, LIAO E Q, et al. Transfer learning for traffic speed prediction: A preliminary study[C]//Proc. Workshops Thirty-Second AAAI Conf. Artificial Intelligence. 2018: 174-177.
[29] MALLICK T, BALAPRAKASH P, RASK E, et al. Transfer Learning with Graph Neural Net- works for Short-Term Highway Traffic Forecasting[J]. 2020 25th International Conference on Pattern Recognition (ICPR), 2021: 10367-10374.
[30] AGAFONOV A. Traffic Flow Prediction Using Graph Convolution Neural Networks[C/OL]// 2020 10th International Conference on Information Science and Technology (ICIST). 2020: 91-95. DOI: 10.1109/ICIST49303.2020.9201971.
[31] JEFFERY D J, RUSSAM K, ROBERTSON D I. Electronic Route Guidance by Autoguide: The Research Background[J]. Traffic engineering and control, 1987, 28: 525-529.
[32] MAKRIDAKIS S, HIBON M. ARMA models and the Box–Jenkins methodology[J]. Journal of Forecasting, 1997, 16(3): 147-163.
[33] DING A, ZHAO X, JIAO L. Traffic flow time series prediction based on statistics learning theory[J]. Proceedings. The IEEE 5th International Conference on Intelligent Transportation Systems, 2002: 727-730.
[34] HOCHREITER S, SCHMIDHUBER J. Long Short-Term Memory[J/OL]. Neural Computation, 1997, 9(8): 1735-1780. DOI: 10.1162/neco.1997.9.8.1735.
[35] PASCANU R, MIKOLOV T, BENGIO Y. On the difficulty of training recurrent neural net- works[C/OL]//DASGUPTA S, MCALLESTER D. Proceedings of Machine Learning Research: volume 28 Proceedings of the 30th International Conference on Machine Learning. Atlanta, Georgia, USA: PMLR, 2013: 1310-1318. https://proceedings.mlr.press/v28/pascanu13.html.
[36] WU Y, TAN H, QIN L, et al. A hybrid deep learning based traffic flow prediction method and its understanding[J]. Transportation Research Part C-emerging Technologies, 2018, 90: 166-180.
[37] ZHANG J, ZHENG Y, QI D. Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction[C]//AAAI. 2017.
[38] YU J J, MARKOS C, ZHANG S. Long-term urban traffic speed prediction with deep learning on graphs[J]. traffic, 2021, 9(20): 21.
[39] ZHANG C, ZHANG S, YU J J Q, et al. FASTGNN: A topological information protected fed- erated learning approach for traffic speed forecasting[J]. IEEE Transactions on Industrial Infor- matics, 2021, 17(12): 8464-8474.
[40] YU J J Q. Graph Construction for Traffic Prediction: A Data-Driven Approach[J]. IEEE Trans- actions on Intelligent Transportation Systems, 2022.
[41] GUO S, LIN Y, FENG N, et al. Attention based spatial-temporal graph convolutional networks for traffic flow forecasting[C]//Proceedings of the AAAI conference on artificial intelligence: volume 33. 2019: 922-929.
[42] DEFFERRARD M, BRESSON X, VANDERGHEYNST P. Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering[J/OL]. CoRR, 2016, abs/1606.09375. http: //arxiv.org/abs/1606.09375.
[43] VELIČKOVIĆ P, CUCURULL G, CASANOVA A, et al. Graph attention networks[A]. 2017.
[44] GAO H, JI S. Graph u-nets[C]//international conference on machine learning. PMLR, 2019: 2083-2092.
[45] YAN S, XIONG Y, LIN D. Spatial temporal graph convolutional networks for skeleton- based action recognition[C]//Proceedings of the AAAI conference on artificial intelligence: vol- ume 32. 2018.
[46] HE Y, ZHAO Y, WANG H, et al. GC-LSTM: A deep spatiotemporal model for passenger flow forecasting of high-speed rail network[C]//2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2020: 1-6.
[47] FRASCA F, ROSSI E, EYNARD D, et al. Sign: Scalable inception graph neural networks[A]. 2020.
[48] WANG L, GENG X, MA X, et al. Cross-city transfer learning for deep spatio-temporal predic- tion[C]//Proc. 28th Int. Joint Conf. Artif. Intell. (IJCAI). S. Kraus, Ed., 2019: 1893-1899.
[49] YAO H, LIU Y, WEI Y, et al. Learning from Multiple Cities: A Meta-Learning Approach for Spatial-Temporal Prediction[J]. The World Wide Web Conference, 2019.
[50] MACQUEEN J, et al. Some methods for classification and analysis of multivariate observations [C]//Proceedings of the fifth Berkeley symposium on mathematical statistics and probability: volume 1. Oakland, CA, USA, 1967: 281-297.
[51] HADI A S. Finding Groups in Data: An Introduction to Chster Analysis[J]. Technometrics, 1991, 34: 111-112.
[52] WANG X, SMITH K, HYNDMAN R. Characteristic-based clustering for time series data[J]. Data mining and knowledge Discovery, 2006, 13(3): 335-364.
[53] ESTER M, KRIEGEL H P, SANDER J, et al. A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise[C]//KDD. 1996.
[54] AGHABOZORGI S R, SHIRKHORSHIDI A S, WAH T Y. Time-series clustering - A decade review[J]. Inf. Syst., 2015, 53: 16-38.
[55] RAMONI M, SEBASTIANI P, COHEN P. Multivariate clustering by dynamics[C]// AAAI/IAAI. 2000: 633-638.
[56] WANG X, SMITH K A, HYNDMAN R, et al. A scalable method for time series clustering[J]. Unrefereed research papers, 2004, 1.
[57] BIERNACKI C, CELEUX G, GOVAERT G. Assessing a mixture model for clustering with the integrated completed likelihood[J]. IEEE transactions on pattern analysis and machine in- telligence, 2000, 22(7): 719-725.
[58] ZHENG Y, CAPRA L, WOLFSON O, et al. Urban Computing: Concepts, Methodologies, and Applications[J]. ACM Transactions on Intelligent Systems and Technology, 2014, 5(3): 38.
[59] WANG L, GUO B, YANG Q. Smart City Development With Urban Transfer Learning[J]. IEEE Computer, 2018, 51(12): 32-41.
[60] WANG L, GENG X, KE J, et al. Ridesourcing car detection by transfer learning[A]. 2017.
[61] YANG D, ZHANG D, QU B. Participatory Cultural Mapping Based on Collective Behavior Data in Location-Based Social Networks[J]. ACM Transactions on Intelligent Systems and Technology, 2016, 7(3): 30.
[62] KARYPIS G, KUMAR V. A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs[J]. SIAM Journal on Scientific Computing, 1998, 20(1): 359-392.
[63] KRISHNAKUMARI P, VAN LINT H, DJUKIC T, et al. A data driven method for OD matrix estimation[J]. Transportation Research Part C-emerging Technologies, 2019, 113: 38-56.
[64] KIPF T N, WELLING M. Semi-Supervised Classification with Graph Convolutional Networks [J/OL]. CoRR, 2016, abs/1609.02907. http://arxiv.org/abs/1609.02907.
[65] NAIR V, HINTON G E. Rectified Linear Units Improve Restricted Boltzmann Machines[C]//Proceedings of the 27th International Conference on Machine Learning. 2010: 807-814.
[66] KINGMA D P, BA J L. Adam: A Method for Stochastic Optimization[C]//ICLR 2015 : Inter- national Conference on Learning Representations 2015. 2015.
[67] HUANG R, HUANG C, LIU Y, et al. LSGCN: Long Short-Term Traffic Prediction with Graph Convolutional Networks.[C]//IJCAI: volume 7. 2020: 2355-2361.
[68] YIN X, WU G, WEI J, et al. Deep learning on traffic prediction: Methods, analysis, and future directions[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 23(6): 4927- 4943.
[69] BAI L, YAO L, LI C, et al. Adaptive Graph Convolutional Recurrent Network for Traffic Fore- casting: abs/2007.02842[A]. 2020.
[70] DU Y, WANG J, FENG W, et al. AdaRNN: Adaptive Learning and Forecasting of Time Se- ries[J]. Proceedings of the 30th ACM International Conference on Information & Knowledge Management, 2021.
[71] GENG X, LI Y, WANG L, et al. Spatiotemporal Multi-Graph Convolution Network for Ride- Hailing Demand Forecasting[C]//AAAI. 2019.
[72] ZOU Y, HUA X, ZHANG Y, et al. Hybrid short-term freeway speed prediction methods based on periodic analysis[J]. Canadian Journal of Civil Engineering, 2015, 42: 570-582.
[73] LINSKER R. Self-organization in a perceptual network[J]. Computer, 1988, 21: 105-117.
[74] HJELM R D, FEDOROV A, LAVOIE-MARCHILDON S, et al. Learning deep representations by mutual information estimation and maximization: abs/1808.06670[A]. 2019.
[75] CHEN T, GUESTRIN C. Xgboost: A scalable tree boosting system[C]//Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. 2016: 785-794.
[76] ZHU Y, XU Y, LIU Q, et al. An empirical study of graph contrastive learning[A]. 2021.
[77] LEDUC G, et al. Road traffic data: Collection methods and applications[J]. Work. Papers Energy, Transp. Climate Change, 2008, 1(55): 1-55.
[78] TEDJOPURNOMO D A, BAO Z, ZHENG B, et al. A survey on modern deep neural network for traffic prediction: Trends, methods and challenges[J]. IEEE Trans. Knowl. Data Eng., 2020,34(4): 1544-1561.
[79] COSTA C, CHATZIMILIOUDIS G, ZEINALIPOUR-YAZTI D, et al. Towards real-time road traffic analytics using telco big data[C]//Proc. Int. workshop on real-time business intelligence and analytics. 2017: 1-5.
[80] NEILSON A, DANIEL B, TJANDRA S, et al. Systematic review of the literature on big data in the transportation domain: Concepts and applications[J]. Big Data Res., 2019, 17: 35-44.
[81] JIANG R, YIN D, WANG Z, et al. Dl-traff: Survey and benchmark of deep learning models for urban traffic prediction[C]//Proc. 30th ACM int. conf. inf. knowl. manage. 2021: 4515-4525.
[82] MANIBARDO E L, LAÑA I, DEL SER J. Deep learning for road traffic forecasting: Does it make a difference?[J]. IEEE Trans. Intell. Transp. Syst., 2021, 23(7): 6164-6188.
[83] HOLMGREN J, FREDRIKSSON H, DAHL M. Traffic data collection using active mobile and stationary devices[J]. Procedia Comput. Sci., 2020, 177: 49-56.
[84] FANG Z, WU D, PAN L, et al. When Transfer Learning Meets Cross-City Urban Flow Prediction: Spatio-Temporal Adaptation Matters[J]. Proc. 22nd Int. Joint Conf. Artif. Intell (IJCAI), 2022: 2030-2036.