AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction
Traffic congestion prediction has become an indispensable component of an intelligent transport system. However, one limitation of the existing methods is that they treat the effects of spatio-temporal correlations on traffic prediction as invariable during modeling spatio-temporal features, which r...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/4/1261 |
| _version_ | 1797297025741488128 |
|---|---|
| author | Lecheng Li Fei Dai Bi Huang Shuai Wang Wanchun Dou Xiaodong Fu |
| author_facet | Lecheng Li Fei Dai Bi Huang Shuai Wang Wanchun Dou Xiaodong Fu |
| author_sort | Lecheng Li |
| collection | DOAJ |
| description | Traffic congestion prediction has become an indispensable component of an intelligent transport system. However, one limitation of the existing methods is that they treat the effects of spatio-temporal correlations on traffic prediction as invariable during modeling spatio-temporal features, which results in inadequate modeling. In this paper, we propose an attention-based spatio-temporal 3D residual neural network, named AST3DRNet, to directly forecast the congestion levels of road networks in a city. AST3DRNet combines a 3D residual network and a self-attention mechanism together to efficiently model the spatial and temporal information of traffic congestion data. Specifically, by stacking 3D residual units and 3D convolution, we proposed a 3D convolution module that can simultaneously capture various spatio-temporal correlations. Furthermore, a novel spatio-temporal attention module is proposed to explicitly model the different contributions of spatio-temporal correlations in both spatial and temporal dimensions through the self-attention mechanism. Extensive experiments are conducted on a real-world traffic congestion dataset in Kunming, and the results demonstrate that AST3DRNet outperforms the baselines in short-term (5/10/15 min) traffic congestion predictions with an average accuracy improvement of 59.05%, 64.69%, and 48.22%, respectively. |
| first_indexed | 2024-03-07T22:15:17Z |
| format | Article |
| id | doaj.art-bbf188e5ea9c4ad2af1d5eb5076aca38 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-07T22:15:17Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-bbf188e5ea9c4ad2af1d5eb5076aca382024-02-23T15:34:00ZengMDPI AGSensors1424-82202024-02-01244126110.3390/s24041261AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion PredictionLecheng Li0Fei Dai1Bi Huang2Shuai Wang3Wanchun Dou4Xiaodong Fu5School of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, ChinaSchool of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, ChinaSchool of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, ChinaSchool of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, ChinaState Key Laboratory for Novel Software Technology, Department of Computer Science and Technology, Nanjing University, Nanjing 210008, ChinaFaculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, ChinaTraffic congestion prediction has become an indispensable component of an intelligent transport system. However, one limitation of the existing methods is that they treat the effects of spatio-temporal correlations on traffic prediction as invariable during modeling spatio-temporal features, which results in inadequate modeling. In this paper, we propose an attention-based spatio-temporal 3D residual neural network, named AST3DRNet, to directly forecast the congestion levels of road networks in a city. AST3DRNet combines a 3D residual network and a self-attention mechanism together to efficiently model the spatial and temporal information of traffic congestion data. Specifically, by stacking 3D residual units and 3D convolution, we proposed a 3D convolution module that can simultaneously capture various spatio-temporal correlations. Furthermore, a novel spatio-temporal attention module is proposed to explicitly model the different contributions of spatio-temporal correlations in both spatial and temporal dimensions through the self-attention mechanism. Extensive experiments are conducted on a real-world traffic congestion dataset in Kunming, and the results demonstrate that AST3DRNet outperforms the baselines in short-term (5/10/15 min) traffic congestion predictions with an average accuracy improvement of 59.05%, 64.69%, and 48.22%, respectively.https://www.mdpi.com/1424-8220/24/4/1261traffic congestion prediction3D convolution3D residual unitself-attention mechanismspatio-temporal attention |
| spellingShingle | Lecheng Li Fei Dai Bi Huang Shuai Wang Wanchun Dou Xiaodong Fu AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction Sensors traffic congestion prediction 3D convolution 3D residual unit self-attention mechanism spatio-temporal attention |
| title | AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction |
| title_full | AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction |
| title_fullStr | AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction |
| title_full_unstemmed | AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction |
| title_short | AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction |
| title_sort | ast3drnet attention based spatio temporal 3d residual neural networks for traffic congestion prediction |
| topic | traffic congestion prediction 3D convolution 3D residual unit self-attention mechanism spatio-temporal attention |
| url | https://www.mdpi.com/1424-8220/24/4/1261 |
| work_keys_str_mv | AT lechengli ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction AT feidai ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction AT bihuang ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction AT shuaiwang ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction AT wanchundou ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction AT xiaodongfu ast3drnetattentionbasedspatiotemporal3dresidualneuralnetworksfortrafficcongestionprediction |