Attention-based Conv-LSTM and Bi-LSTM networks for large-scale traffic speed prediction

被引:0
|
作者
Xiaojian Hu
Tong Liu
Xiatong Hao
Chenxi Lin
机构
[1] Southeast University,Jiangsu Key Laboratory of Urban ITS
[2] Southeast University,Jiangsu Province Collaborative Innovation Center of Modern Urban Traffic Technologies
[3] National Demonstration Center for Experimental Road and Traffic Engineering Education (Southeast University),School of Transportation
[4] Southeast University,undefined
来源
关键词
Traffic speed prediction; Conv-LSTM; Bi-LSTM; Attention mechanism; Spatiotemporal; Periodic;
D O I
暂无
中图分类号
学科分类号
摘要
Timely and accurate traffic speed prediction has gained increasing importance for urban traffic management and helping one to make advisable travel decision. However, the existing approaches have difficulty extracting features of large-scale traffic data. This study proposed a hybrid deep learning method named AB-ConvLSTM for large-scale traffic speed prediction. The proposed model consists of a convolutional-long short-term memory (Conv-LSTM) module, an attention mechanism module, and two bidirectional LSTM (Bi-LSTM) modules. Conv-LSTM networks are used to extract the spatiotemporal features of traffic speed data. In addition, the attention mechanism module is introduced to enhance the performance of Conv-LSTM by automatically capturing the importance of different historical periods to the final prediction and assigning corresponding weights. What's more, two Bi-LSTM networks are designed to extract daily and weekly periodic features and capture variation tendency from forward and backward traffic data. Experimental results carried out on urban road networks show that the proposed model consistently outperforms the competing models.
引用
收藏
页码:12686 / 12709
页数:23
相关论文
共 50 条
  • [1] Attention-based Conv-LSTM and Bi-LSTM networks for large-scale traffic speed prediction
    Hu, Xiaojian
    Liu, Tong
    Hao, Xiatong
    Lin, Chenxi
    [J]. JOURNAL OF SUPERCOMPUTING, 2022, 78 (10): : 12686 - 12709
  • [2] A Hybrid Deep Learning Model With Attention-Based Conv-LSTM Networks for Short-Term Traffic Flow Prediction
    Zheng, Haifeng
    Lin, Feng
    Feng, Xinxin
    Chen, Youjia
    [J]. IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2021, 22 (11) : 6910 - 6920
  • [3] Attention-based Bi-LSTM Model for Anomalous HTTP Traffic Detection
    Yu, Yuqi
    Liu, Guannan
    Yan, Hanbing
    Li, Hong
    Guan, Hongchao
    [J]. 2018 15TH INTERNATIONAL CONFERENCE ON SERVICE SYSTEMS AND SERVICE MANAGEMENT (ICSSSM), 2018,
  • [4] Attention Based Graph Bi-LSTM Networks for Traffic Forecasting
    Zhao, Han
    Yang, Huan
    Wang, Yu
    Wang, Danwei
    Su, Rong
    [J]. 2020 IEEE 23RD INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS (ITSC), 2020,
  • [5] Short-Term Traffic Flow Prediction with Conv-LSTM
    Liu, Yipeng
    Zheng, Haifeng
    Feng, Xinxin
    Chen, Zhonghui
    [J]. 2017 9TH INTERNATIONAL CONFERENCE ON WIRELESS COMMUNICATIONS AND SIGNAL PROCESSING (WCSP), 2017,
  • [6] Attention-Based Bi-LSTM Network for Abusive Language Detection
    Nelatoori, Kiran Babu
    Kommanti, Hima Bindu
    [J]. IETE JOURNAL OF RESEARCH, 2023, 69 (11) : 7884 - 7892
  • [7] Attention-Based Bi-LSTM Model for Arabic Depression Classification
    Almars, Abdulqader M.
    [J]. CMC-COMPUTERS MATERIALS & CONTINUA, 2022, 71 (02): : 3091 - 3106
  • [8] Attention-Based Bi-LSTM for Chinese Named Entity Recognition
    Zhang, Kai
    Ren, Weiping
    Zhang, Yangsen
    [J]. CHINESE LEXICAL SEMANTICS, CLSW 2018, 2018, 11173 : 643 - 652
  • [9] Real-time risk prediction of chemical processes based on attention-based Bi-LSTM
    Qianlin Wang
    Jiaqi Han
    Feng Chen
    Xin Zhang
    Cheng Yun
    Zhan Dou
    Tingjun Yan
    Guoan Yang
    [J]. Chinese Journal of Chemical Engineering, 2024, 75 (11) - 141
  • [10] Sequence to sequence hybrid Bi-LSTM model for traffic speed prediction
    Ounoughi, Chahinez
    Ben Yahia, Sadok
    [J]. EXPERT SYSTEMS WITH APPLICATIONS, 2024, 236