The invention relates to a GraphSAGE-GAN-based traffic network data restoration method, which comprises the following steps of calculating the correlation between road network detectors according to the historical data of the detectors in a road network to obtain a road network correlation matrix, and constructing a road network structure based on time correlation according to the obtained road network correlation matrix, secondly, extracting potential spatial-temporal features for constructing a road network structure by utilizing Graph-SAGE, and finally, taking the extracted spatial-temporalfeatures as input of a generator in the generative adversarial network, so that the generator can generate complete road network traffic state information according to the spatial-temporal features extracted by the GraphSAGE through adversarial training of the generative adversarial network, and therefore, realizing restoration of the road network traffic state data. Spatial features between theroad network traffic state detectors can be deeply excavated, and the precision of road network traffic state restoration is effectively improved.
一种基于GraphSAGE‑GAN的交通路网数据修复方法,首先根据路网中检测器的历史数据计算路网检测器之间的相关性,得到路网相关性矩阵,然后根据得到的路网相关性矩阵构建基于时间相关性的路网结构。其次,利用GraphSAGE提取构建路网结构的潜在时空特征,最后将提取的时空特征作为生成对抗网络中生成器的输入,使其通过生成对抗网络的对抗训练,生成器能够根据经GraphSAGE提取的时空特征生成完整的的路网交通状态信息,从而实现路网交通状态数据的修复。本发明可以深度挖掘路网交通状态检测器之间的空间特征,有效提高路网交通状态修复的精度。
GraphSAGE-GAN-based traffic network data restoration method
一种基于GraphSAGE-GAN的交通路网数据修复方法
2021-02-02
Patent
Electronic Resource
Chinese
Intelligent network connection traffic data restoration method
European Patent Office | 2023
|Road network traffic data restoration method based on DTW-RGCN
European Patent Office | 2020
|