The invention discloses a traffic state estimation method based on floating car data weighted tensor reconstruction. The method comprises the following steps: acquiring floating car data; constructinga tensor model and a weight factor tensor model according to the floating car data, the tensor model and the weight factor tensor model having the same size and order, and each initial assignment ofthe weight factor tensor being 1; designing a tensor reconstruction algorithm introducing a weight factor; re-assigning the weight factor tensor according to each factor influencing the reliability ofthe floating car data; and obtaining floating car data and traffic state information according to the re-assigned weight factor tensor, the floating vehicle data tensor model and the tensor reconstruction algorithm. According to the method, the constructed tensor reconstruction model can better meet the requirement of road network traffic state estimation, so that more accurate road network traffic state estimation is obtained.
本发明公开了一种基于浮动车数据加权张量重建的交通状态估计方法,所述方法包括:获取浮动车数据;根据所述浮动车数据构建张量模型和权重因子张量模型,所述张量模型和所述权重因子张量模型具有相同尺寸和阶数,权重因子张量的各初始赋值为1;设计引入权重因子的张量重建算法;根据各个影响浮动车数据可靠性的因素对权重因子张量进行重新赋值;根据所述重新赋值后的权重因子张量、所述浮动车数据张量模型与所述张量重建算法,获得浮动车数据和交通状态信息。本发明可以使得构建的张量重建模型更符合道路网交通状态估计的需求,从而获得更精确的道路网交通状态估计。
Traffic state estimation method based on floating car data weighted tensor reconstruction
一种基于浮动车数据加权张量重建的交通状态估计方法
2020-06-19
Patent
Electronic Resource
Chinese
A Tensor Completion-Based Traffic State Estimation Model
ASCE | 2014
|Dynamic Traffic Information Estimation Based on Floating Car Data
Springer Verlag | 2018
|Full-region road section traffic flow estimation method based on floating car data
European Patent Office | 2024
|