计算机科学 ›› 2025, Vol. 52 ›› Issue (6A): 240400149-8.doi: 10.11896/jsjkx.240400149
郑创锐, 邓秀勤, 陈磊
ZHENG Chuangrui, DENG Xiuqin, CHEN Lei
摘要： 交通预测对城市规划和交通管理起着关键作用,基于机器学习和统计学的传统预测方法在捕捉复杂的非线性关系和长期依赖方面的能力有限,无法捕获交通路网中复杂的时空关系。现有基于图神经网络(GNN)的预测模型大多采用预定的静态图,无法准确反映真实道路网络的拓扑结构,且大多数模型只简单地考虑交通流量在不同节点之间的传播过程,忽略了每个节点本身流量的生成过程。针对以上问题,提出了一种解耦自适应动态图卷积网络模型(Decoupled Adaptive Dynamic Graph Convolutional Network,DADGCN),该模型通过一个自适应动态图模块,有效地量化不同节点间的动态相关性,从而捕捉交通网络中复杂的空间依赖关系,同时通过数据驱动的方式将节点的流量解耦为传播流量和生成流量,利用多头自注意力机制来处理解耦后的信号,从而提高了模型处理复杂交通数据的灵活性,提升了预测精度。实验结果表明,在数据集METR-LA上,DADGCN在60 min上的MAE比基于扩散卷积的模型DCRNN和Graph Wavenet分别提升了7.78%,10.14%;在数据集PEMS-BAY上DADGCN分别提升了25.39%,21.19%。在数据集PEMS04和PEMS08上,DADGCN比基于自适应图模型MTGNN在MAPE和RMSE上分别提升了11.61%和3.90%,表明该模型不仅能够更深入地理解交通流中的固有动态特征,还能够适应各种复杂环境下的变化,为城市交通管理和规划提供更准确、更可靠的数据支持。
中图分类号:
| [1]KUMAR S V,VANAJAKSHI L.Short-term traffic flow prediction using seasonal ARIMA model with limited input data[J].European Transport Research Review,2015,7:1-9. [2]SCHIMBINSCHI F,MOREIRA-MATIAS L,NGUYEN V X,et al.Topology-regularized universal vector autoregression for traffic forecasting in large urban areas[J].Expert Systems with Applications,2017,82:301-316. [3]YU H,JI N,REN Y,et al.A special event-based K-nearest neighbor model for short-term traffic state prediction[J].IEEE Access,2019,7:81717-81729. [4]FENG X,LING X,ZHENG H,et al.Adaptive multi-kernelSVM with spatial-temporal correlation for short-term traffic flow prediction[J].IEEE Transactions on Intelligent Transportation Systems,2018,20(6):2001-2013. [5]SHAO W,JIN Z,WANG S,et al.Long-term Spatio-Temporal Forecasting via Dynamic Multiple-Graph Attention[C]//31st International Joint Conference on Artificial Intelligence(IJCAI 2022).International Joint Conferences on Artificial Intelligence,2022:2225-2232. [6]FENG A,TASSIULA L.Adaptive Graph Spatial-TemporalTransformer Network for Traffic Forecasting[C]//Proceedings of the 31st ACM International Conference on Information & Knowledge Management.2022:3933-3937. [7]LUO X,ZHU C,ZHANG D,et al.Dynamic Graph Convolutional Network with Attention Fusion for Traffic Flow Prediction[M]//ECAI 2023.IOS Press,2023:1633-1640. [8]LI Y,YU R,SHAHABI C,et al.Diffusion Convolutional Recurrent Neural Network:Data-Driven Traffic Forecasting[J].arXiv:1707.01926. [9]HAN H,ZHANG M,HOU M,et al.STGCN:a spatial-temporal aware graph learning method for POI recommendation[C]//International Conference on Data Mining(ICDM).IEEE,2020:1052-1057. [10]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.2019,33(1):922-929. [11]XIAO Y,QIN J Y,LI K L,et al.Multimodal Traffic flow colla-borative predictionMethod based on graph contrast learning [J].Computer Science,2023,50(10):135-145. [12]HE Z,ZHAO C,HUANG Y.Multivariate time series deep spatiotemporal forecasting with graph neural network[J].Applied Sciences,2022,12(11):5731. [13]UBAL C,DI-GIORG G,CONTRERAS-REYES J E,et al.Predicting the Long-Term Dependencies in Time Series Using Recurrent Artificial Neural Networks[J].Machine Learning and Knowledge Extraction,2023,5(4):1340-1358. [14]MUMUNI A,MUMUNI F.Data augmentation:A comprehen-sive survey of modern approaches[J].Array,2022,16:100258. [15]LI Z,HU Z,LUO W,et al.SaberNet:Self-attention based effective relation network for few-shot learning[J].Pattern Recognition,2023,133:109024. [16]QIN C,HUANG G,YU H,et al.Geological information prediction for shield machine usingan enhanced multi-head self-attention convolution neural network with two-stage feature extraction[J].Geoscience Frontiers,2023,14(2):101519. [17]ZHANG S,TONG H,XU J,et al.Graph convolutional net-works:a comprehensive review[J].Computational Social Networks,2019,6(1):1-23. [18]SUTSKEVER I,VINYALS O,LE Q V.Sequence to sequence learning with neural networks[J].arXiv:1409.3215. [19]WU Z,PAN S,LONG G,et al.Graph wavenet for deep spatial-temporal graph modeling[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence.2019:1907-1913. [20]YU B,YIN H,ZHU Z.Spatio-temporal graph convolutional networks:A deep learning framework for traffic forecasting[J].arXiv:1709.04875,2017. [21]WU Z,PAN S,LONG G,et al.Connecting the dots:Multivariate time series forecasting with graph neural networks[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.2020:753-763. [22]TA X,LIU Z,HU X,et al.Adaptive spatio-temporal graph neural network for traffic forecasting[J].Knowledge-based Systems,2022,242:108199. [23]SUN Y,JIANG X,HU Y,et al.Dual dynamic spatial-temporal graph convolution network for traffic prediction[J].IEEE Transactions on Intelligent Transportation Systems,2022,23(12):23680-23693. [24]LI F,FENG J,YAN H,et al.Dynamic graph convolutional recurrent network for traffic prediction:Benchmark and solution[J].ACM Transactions on Knowledge Discovery from Data,2023,17(1):1-21. [25]WU B,CHEN L.DSTCGCN:Learning Dynamic Spatial-Temporal Cross Dependencies for Traffic Forecasting[J].arXiv:2307.00518,2023. |
|
||
首页