Vehicle Interaction Behavior Prediction with Self-Attention
The structured road is a scene with high interaction between vehicles, but due to the high uncertainty of behavior, the prediction of vehicle interaction behavior is still a challenge. This prediction is significant for controlling the ego-vehicle. We propose an interaction behavior prediction model...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-01-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/22/2/429 |
_version_ | 1797490617316540416 |
---|---|
author | Linhui Li Xin Sui Jing Lian Fengning Yu Yafu Zhou |
author_facet | Linhui Li Xin Sui Jing Lian Fengning Yu Yafu Zhou |
author_sort | Linhui Li |
collection | DOAJ |
description | The structured road is a scene with high interaction between vehicles, but due to the high uncertainty of behavior, the prediction of vehicle interaction behavior is still a challenge. This prediction is significant for controlling the ego-vehicle. We propose an interaction behavior prediction model based on vehicle cluster (VC) by self-attention (VC-Attention) to improve the prediction performance. Firstly, a five-vehicle based cluster structure is designed to extract the interactive features between ego-vehicle and target vehicle, such as Deceleration Rate to Avoid a Crash (DRAC) and the lane gap. In addition, the proposed model utilizes the sliding window algorithm to extract VC behavior information. Then the temporal characteristics of the three interactive features mentioned above will be caught by two layers of self-attention encoder with six heads respectively. Finally, target vehicle’s future behavior will be predicted by a sub-network consists of a fully connected layer and SoftMax module. The experimental results show that this method has achieved accuracy, precision, recall, and F1 score of more than 92% and time to event of 2.9 s on a Next Generation Simulation (NGSIM) dataset. It accurately predicts the interactive behaviors in class-imbalance prediction and adapts to various driving scenarios. |
first_indexed | 2024-03-10T00:35:33Z |
format | Article |
id | doaj.art-9b53c3db023b4c428b4d0ea01e3f86fb |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-10T00:35:33Z |
publishDate | 2022-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-9b53c3db023b4c428b4d0ea01e3f86fb2023-11-23T15:18:29ZengMDPI AGSensors1424-82202022-01-0122242910.3390/s22020429Vehicle Interaction Behavior Prediction with Self-AttentionLinhui Li0Xin Sui1Jing Lian2Fengning Yu3Yafu Zhou4Key Laboratory of Energy Conservation and New Energy Vehicle Power Control and Vehicle Technology, School of Automotive Engineering, Faculty of Vehicle Engineering and Mechanics, Dalian University of Technology, Dalian 116081, ChinaKey Laboratory of Energy Conservation and New Energy Vehicle Power Control and Vehicle Technology, School of Automotive Engineering, Faculty of Vehicle Engineering and Mechanics, Dalian University of Technology, Dalian 116081, ChinaKey Laboratory of Energy Conservation and New Energy Vehicle Power Control and Vehicle Technology, School of Automotive Engineering, Faculty of Vehicle Engineering and Mechanics, Dalian University of Technology, Dalian 116081, ChinaKey Laboratory of Energy Conservation and New Energy Vehicle Power Control and Vehicle Technology, School of Automotive Engineering, Faculty of Vehicle Engineering and Mechanics, Dalian University of Technology, Dalian 116081, ChinaKey Laboratory of Energy Conservation and New Energy Vehicle Power Control and Vehicle Technology, School of Automotive Engineering, Faculty of Vehicle Engineering and Mechanics, Dalian University of Technology, Dalian 116081, ChinaThe structured road is a scene with high interaction between vehicles, but due to the high uncertainty of behavior, the prediction of vehicle interaction behavior is still a challenge. This prediction is significant for controlling the ego-vehicle. We propose an interaction behavior prediction model based on vehicle cluster (VC) by self-attention (VC-Attention) to improve the prediction performance. Firstly, a five-vehicle based cluster structure is designed to extract the interactive features between ego-vehicle and target vehicle, such as Deceleration Rate to Avoid a Crash (DRAC) and the lane gap. In addition, the proposed model utilizes the sliding window algorithm to extract VC behavior information. Then the temporal characteristics of the three interactive features mentioned above will be caught by two layers of self-attention encoder with six heads respectively. Finally, target vehicle’s future behavior will be predicted by a sub-network consists of a fully connected layer and SoftMax module. The experimental results show that this method has achieved accuracy, precision, recall, and F1 score of more than 92% and time to event of 2.9 s on a Next Generation Simulation (NGSIM) dataset. It accurately predicts the interactive behaviors in class-imbalance prediction and adapts to various driving scenarios.https://www.mdpi.com/1424-8220/22/2/429vehicle interaction behavior predictionself-attentionvehicle clusterend-to-end predictionclass imbalance |
spellingShingle | Linhui Li Xin Sui Jing Lian Fengning Yu Yafu Zhou Vehicle Interaction Behavior Prediction with Self-Attention Sensors vehicle interaction behavior prediction self-attention vehicle cluster end-to-end prediction class imbalance |
title | Vehicle Interaction Behavior Prediction with Self-Attention |
title_full | Vehicle Interaction Behavior Prediction with Self-Attention |
title_fullStr | Vehicle Interaction Behavior Prediction with Self-Attention |
title_full_unstemmed | Vehicle Interaction Behavior Prediction with Self-Attention |
title_short | Vehicle Interaction Behavior Prediction with Self-Attention |
title_sort | vehicle interaction behavior prediction with self attention |
topic | vehicle interaction behavior prediction self-attention vehicle cluster end-to-end prediction class imbalance |
url | https://www.mdpi.com/1424-8220/22/2/429 |
work_keys_str_mv | AT linhuili vehicleinteractionbehaviorpredictionwithselfattention AT xinsui vehicleinteractionbehaviorpredictionwithselfattention AT jinglian vehicleinteractionbehaviorpredictionwithselfattention AT fengningyu vehicleinteractionbehaviorpredictionwithselfattention AT yafuzhou vehicleinteractionbehaviorpredictionwithselfattention |