基于HDNNF-CAF的短时交通流预测研究

doi:10.3778/j.issn.1002-8331.2411-0084

摘要/Abstract

摘要： 短时交通流预测在智能交通系统中扮演重要的角色。针对交通流复杂多变的时空特征、非平稳性及外部因素引发的数据异常，提出考虑异常因素的混合深度神经网络预测模型（hybrid deep neural network forecasting model considering anomalous factors，HDNNF-CAF）。该模型将邻接矩阵、交通流量矩阵及交通流其他参数矩阵结合异常数据处理理论，进行数据预处理和异常数据识别。建立异常数据时空特征提取理论，捕获异常数据时空信息；利用变分模态分解（VMD）降低交通流数据非平稳性，并提出图卷积网络（GCN）优化Informer理论分别对各个子序列进行特征提取，以组合生成交通流时空信息。最终结合异常数据与交通流数据的时空信息生成预测结果。在真实数据集PeMS04上进行验证，实验结果表明，HDNNF-CAF能够有效识别交通流异常数据，提高预测精度，优于一些现有方法。

关键词: 短时交通流, 预测, 深度学习, 图卷积网络, 时空信息

Abstract: Short-term traffic flow prediction is essential in intelligent transportation systems. To address the spatiotemporal complexity, non-stationarity, and anomalies caused by external factors in traffic flow data, a hybrid deep neural network forecasting model considering anomalous factors (HDNNF-CAF) is proposed. This model combines adjacency matrices, traffic flow matrices, and other traffic parameters with anomaly detection techniques for data preprocessing and anomaly identification. It establishes a spatiotemporal feature extraction framework to capture the characteristics of anomalous data. Variational mode decomposition (VMD) is applied to mitigate data non-stationarity, and a graph convolutional network (GCN) is integrated with the Informer model to extract features from decomposed sub-sequences, constructing comprehensive spatiotemporal traffic information. The final prediction results are obtained by merging the spatiotemporal features of anomalous and normal traffic data. Validation on the PeMS04 dataset shows that HDNNF-CAF effectively identifies anomalies and improves prediction accuracy, outperforming existing methods.

Key words: short-term traffic flow, prediction, deep learning, graph convolutional network, spatiotemporal Information

王庆荣, 慕壮壮, 朱昌锋, 何润田. 基于HDNNF-CAF的短时交通流预测研究[J]. 计算机工程与应用, 2025, 61(15): 318-328.

WANG Qingrong, MU Zhuangzhuang, ZHU Changfeng, HE Runtian. Research on Short-Term Traffic Flow Prediction Based on HDNNF-CAF[J]. Computer Engineering and Applications, 2025, 61(15): 318-328.

参考文献

[1] ZHENG H F, LIN F, FENG X X, et al. A hybrid deep learning model with attention-based conv-LSTM networks for short-term traffic flow prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(11): 6910-6920.
[2] YE J X, ZHAO J J, YE K J, et al. How to build a graph-based deep learning architecture in traffic domain: a survey[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(5): 3904-3924.
[3] WU P, HUANG Z L, PIAN Y Z, et al. A combined deep learning method with attention-based LSTM model for short-term traffic speed forecasting[J]. Journal of Advanced Transportation, 2020, 2020(1): 8863724.
[4] ROCCO DI TORREPADULA F, NAPOLITANO E V, DI MARTINO S, et al. Machine learning for public transportation demand prediction: a systematic literature review[J]. Enginee-ring Applications of Artificial Intelligence, 2024, 137: 109166.
[5] YUAN H T, LI G L. A survey of traffic prediction: from spatio-temporal data to intelligent transportation[J]. Data Science and Engineering, 2021, 6(1): 63-85.
[6] CAI W, YANG J, YU Y, et al. PSO-ELM: a hybrid learning model for short-term traffic flow forecasting[J]. IEEE Access, 2020, 8: 6505-6514.
[7] LI L J, YANG W X, BAI S S, et al. KNN-GNN: a powerful graph neural network enhanced by aggregating K-nearest neighbors in common subspace[J]. Expert Systems with Applications, 2024, 253: 124217.
[8] LIN G C, LIN A J, GU D L. Using support vector regression and K-nearest neighbors for short-term traffic flow prediction based on maximal information coefficient[J]. Information Sciences, 2022, 608: 517-531.
[9] ZHENG X Y, SHAO H D, YAN S, et al. Multiscale information enhanced spatial-temporal graph convolutional network for multivariate traffic flow forecasting via magnifying perceptual scope[J]. Engineering Applications of Artificial Intelligence, 2024, 136: 109010.
[10] FANG W W, ZHUO W H, SONG Y Y, et al. Δ free-LSTM: an error distribution free deep learning for short-term traffic flow forecasting[J]. Neurocomputing, 2023, 526: 180-190.
[11] HUANG B H, BAI Y. Traffic flow prediction based on BRNN[C]//Proceedings of the 2019 IEEE 9th International Conference on Electronics Information and Emergency Communication. Piscataway: IEEE, 2019: 320-323.
[12] LIU S, LIN W T, WANG Y, et al. Convolutional neural network-based bidirectional gated recurrent unit-additive attention mechanism hybrid deep neural networks for short-term traffic flow prediction[J]. Sustainability, 2024, 16(5): 1986.
[13] GENG Z L, XU J, WU R S, et al. STGAFormer: spatial-temporal gated attention transformer based graph neural network for traffic flow forecasting[J]. Information Fusion, 2024, 105: 102228.
[14] YU B, YIN H, ZHU Z. Spatio-temporal graph convolutional networks: a deep learning framework for traffic forecasting[J]. arXiv:1709.04875, 2017.
[15] BAO Y X, SHEN Q Q, CAO Y, et al. Residual attention enhanced time-varying multi-factor graph convolutional network for traffic flow prediction[J]. Engineering Applications of Artificial Intelligence, 2024, 133: 108135.
[16] ZHANG H, ZOU Y J, YANG X X, et al. A temporal fusion transformer for short-term freeway traffic speed multistep prediction[J]. Neurocomputing, 2022, 500: 329-340.
[17] ZHOU H Y, ZHANG S H, PENG J Q, et al. Informer: beyond efficient transformer for long sequence time-series forecasting[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2021: 11106-11115.
[18] SONG Y F, LUO R K, ZHOU T C, et al. Graph attention informer for long-term traffic flow prediction under the impact of sports events[J]. Sensors, 2024, 24(15): 4796.
[19] YU Y, SHANG Q, XIE T. A hybrid model for short-term traffic flow prediction based on variational mode decomposition, wavelet threshold denoising, and long short-term memory neural network[J]. Complexity, 2021, 2021(1): 7756299.
[20] 龚循强, 邱万锦, 吕开云, 等. 变分模态分解与自适应图卷积门控循环网络的交通流量组合预测模型[J]. 武汉大学学报(信息科学版), 2024, 49(12): 2329-2341.
GONG X Q, QIU W J, Lü K Y, et al. A combined traffic flow prediction model based on variational mode decomposition and adaptive graph convolutional gated recurrent network[J]. Geomatics and Information Science of Wuhan University, 2024, 49(12): 2329-2341.
[21] HUANG X H, JIANG Y, TANG J. MAPredRNN: multi-attention predictive RNN for traffic flow prediction by dynamic spatio-temporal data fusion[J]. Applied Intelligence, 2023, 53(16): 19372-19383.
[22] YE J H, XUE S J, JIANG A W. Attention-based spatio-temporal graph convolutional network considering external factors for multi-step traffic flow prediction[J]. Digital Communic-ations and Networks, 2022, 8(3): 343-350.