Low-Rank Hankel Tensor Completion for Traffic Speed Estimation

被引:12
|
作者
Wang, Xudong [1 ]
Wu, Yuankai [1 ]
Zhuang, Dingyi [1 ]
Sun, Lijun [1 ]
机构
[1] McGill Univ, Dept Civil Engn, Montreal, PQ H3A 0C3, Canada
基金
加拿大创新基金会; 加拿大自然科学与工程研究理事会;
关键词
Spatio-temporal traffic data; traffic state esti-mation; missing data imputation; low-rank tensor completion; delay embedding transform; STATE ESTIMATION; MATRIX COMPLETION; FILTER; FLOW;
D O I
10.1109/TITS.2023.3247961
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
This paper studies the traffic state estimation (TSE) problem using sparse observations from mobile sensors. Most existing TSE methods either rely on well-defined physical traffic flow models or require large amounts of simulation data as input to train machine learning models. Different from previous studies, we propose a purely data-driven and model-free solution in this paper. We consider the TSE as a spatiotemporal matrix completion/interpolation problem, and apply spatiotemporal delay embedding to transform the original incomplete matrix into a fourth-order Hankel structured tensor. By imposing a low-rank assumption on this tensor structure, we can approximate and characterize both global and local spatiotemporal patterns in a data-driven manner. We use the truncated nuclear norm of a balanced spatiotemporal unfolding -- in which each column represents the vectorization of a small patch in the original matrix -- to approximate the tensor rank. An efficient solution algorithm based on the Alternating Direction Method of Multipliers (ADMM) is developed for model learning. The proposed framework only involves two hyperparameters, spatial and temporal window lengths, which are easy to set given the degree of data sparsity. We conduct numerical experiments on real-world high-resolution trajectory data, and our results demonstrate the effectiveness and superiority of the proposed model in some challenging scenarios.
引用
收藏
页码:4862 / 4871
页数:10
相关论文
共 50 条
  • [41] Dictionary Learning With Low-Rank Coding Coefficients for Tensor Completion
    Jiang, Tai-Xiang
    Zhao, Xi-Le
    Zhang, Hao
    Ng, Michael K.
    [J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 34 (02) : 932 - 946
  • [42] Tensor Completion Via Collaborative Sparse and Low-Rank Transforms
    Li, Ben-Zheng
    Zhao, Xi-Le
    Wang, Jian-Li
    Chen, Yong
    Jiang, Tai-Xiang
    Liu, Jun
    [J]. IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING, 2021, 7 : 1289 - 1303
  • [43] Accurate Tensor Completion via Adaptive Low-Rank Representation
    Zhang, Lei
    Wei, Wei
    Shi, Qinfeng
    Shen, Chunhua
    van den Hengel, Anton
    Zhang, Yanning
    [J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2020, 31 (10) : 4170 - 4184
  • [44] Unifying tensor factorization and tensor nuclear norm approaches for low-rank tensor completion
    Du, Shiqiang
    Xiao, Qingjiang
    Shi, Yuqing
    Cucchiara, Rita
    Ma, Yide
    [J]. NEUROCOMPUTING, 2021, 458 : 204 - 218
  • [45] Low-Rank Tensor Completion Pansharpening Based on Haze Correction
    Wang, Peng
    Su, Yiyang
    Huang, Bo
    Zhu, Daiyin
    Liu, Wenjian
    Nedzved, Alexander
    Krasnoproshin, Viktor V.
    Leung, Henry
    [J]. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62 : 1 - 20
  • [46] On Polynomial Time Methods for Exact Low-Rank Tensor Completion
    Dong Xia
    Ming Yuan
    [J]. Foundations of Computational Mathematics, 2019, 19 : 1265 - 1313
  • [47] Low-rank tensor completion with sparse regularization in a transformed domain
    Wang, Ping-Ping
    Li, Liang
    Cheng, Guang-Hui
    [J]. NUMERICAL LINEAR ALGEBRA WITH APPLICATIONS, 2021, 28 (06)
  • [48] Low-rank tensor completion via smooth matrix factorization
    Zheng, Yu-Bang
    Huang, Ting-Zhu
    Ji, Teng-Yu
    Zhao, Xi-Le
    Jiang, Tai-Xiang
    Ma, Tian-Hui
    [J]. APPLIED MATHEMATICAL MODELLING, 2019, 70 : 677 - 695
  • [49] Riemannian conjugate gradient method for low-rank tensor completion
    Duan, Shan-Qi
    Duan, Xue-Feng
    Li, Chun-Mei
    Li, Jiao-Fen
    [J]. ADVANCES IN COMPUTATIONAL MATHEMATICS, 2023, 49 (03)
  • [50] Low-Rank Tensor Completion with Spatio-Temporal Consistency
    Wang, Hua
    Nie, Feiping
    Huang, Heng
    [J]. PROCEEDINGS OF THE TWENTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2014, : 2846 - 2852