摘要
传统交通流量预测模型未能有效考虑地区间和时段间的时间延迟效应,且难以同时捕捉交通流量的短期波动与长期趋势。为此,提出一种结合小波分解与时间延迟感知的时空预测模型(WTA-LAGNN)。首先,结合小波分解将交通流量数据分为长期趋势模式和短期波动模式:短期波动模式通过特征增强模块强化关键特征,提升对短期波动的敏感性;针对长期趋势,设计了序列增强的多头自注意力机制捕捉流量的长期变化。其次,为了处理时间延迟效应,设计了时间序列延迟感知层,优化区域间流量传播的时空依赖关系。最后,通过融合层生成最终预测结果。基于现实高速公路交通数据集PeMS03、PeMS04、PeMS07、PeMS08进行60 min流量预测,结果表明,在PeMS03和PeMS07数据集上,与时空图神经控制微分方程(STG-NCDE)相比,WTA-LAGNN的平均绝对误差(MAE)、均方根误差(RMSE)分别降低了5.14%、2.69%和5.80%、2.69%;在PeMS08数据集上,与交通流量矩阵-图卷积注意力模型(TFM-GCAM)相比,WTA-LAGNN的MAE、RMSE分别下降了9.28%、3.32%;在PeMS04数据集上,与时空融合图卷积网络(STFGCN)相比,WTA-LAGNN的MAE、RMSE分别降低了3.53%、2.72%。WTA-LAGNN的整体模型性能上优于对比模型,能更有效地捕捉时空依赖关系,提升流量预测精度。
Traditional traffic flow prediction models often fail to effectively account for temporal delays across regions and time periods,and struggle to capture both short-term fluctuations and long-term trends in traffic flow.To address these limitations,a Wavelet Transform and Attention-based Latency-Aware long short Graph Neural Network(WTA-LAGNN)was proposed.Firstly,wavelet decomposition was applied to separate traffic flow data into long-term trend and short-term fluctuation patterns.Key features in the short-term fluctuation pattern were enhanced using a feature enhancement module,improving the models sensitivity to short-term variations.For the long-term trend,a sequence-enhanced multi-head self-attention was designed to capture sustained changes in flow.To address temporal delay effects,a time delay-aware layer was designed to optimize spatio-temporal dependencies in traffic flow propagation between regions.Finally,the fusion layer outputted the prediction results.Experiments were conducted on real highway traffic datasets including PeMS03,PeMS04,PeMS07 and PeMS08 for 60-minute flow prediction.The results showed that compared to Spatio-Temporal Graph Neural Controlled Differential Equation(STG-NCDE),WTA-LAGNN reduced the Mean Absolute Error(MAE)and Root Mean Square Error(RMSE)by 5.14%and 2.69%,as well as 5.80%and 2.69%on the PeMS03 and PeMS07 datasets,respectively;compared to Traffic Flow Matrix-based Graph Convolutional Attention Mechanism(TFM-GCAM),WTA-LAGNN reduced the MAE and RMSE by 9.28%and 3.32%on PeMS08;compared to Spatio-Temporal Fusion Graph Convolutional Network(STFGCN),WTA-LAGNN reduced the MAE and RMSE by 3.53%and 2.72%on PeMS04,respectively.These results demonstrate that WTA-LAGNN outperforms comparison baseline models,and can effectively capture spatio-temporal dependencies,thereby improving traffic flow prediction precision.
作者
潘理虎
张梦麟
樊光瑞
张林梁
张睿
PAN Lihu;ZHANG Menglin;FAN Guangrui;ZHANG Linliang;ZHANG Rui(School of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan Shanxi 030024,China;Shanxi Intelligent Transportation Institute Company Limited,Taiyuan Shanxi 030036,China)
出处
《计算机应用》
北大核心
2025年第11期3649-3657,共9页
journal of Computer Applications
基金
山西省基础研究计划项目(202203021221145)
山西省研究生联合培养示范基地项目(2022JD11)。
关键词
交通流量预测
小波分解
时间延迟感知
注意力机制
时空模型
traffic flow prediction
wavelet decomposition
time delay awareness
attention mechanism
spatio-temporal model
作者简介
通信作者:潘理虎(1974-),男,河南上蔡人,教授,博士,CCF高级会员,主要研究方向:人工智能、智能软件工程,电子邮箱:panlh@tyust.edu.n;张梦麟(2001-),男,山西太原人,硕士研究生,主要研究方向:交通预测、人工智能;樊光瑞(1992-),男,山西汾阳人,博士研究生,主要研究方向:深度学习、智慧交通、社交网络、大数据;张林梁(1984-),男,山东安丘人,高级工程师,博士,主要研究方向:深度学习、机器学习、计算机视觉、大数据挖掘;张睿(1987-),男,山西太原人,副教授,博士,CCF高级会员,主要研究方向:智能信息处理。
