Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

Complementary antioxidant strategy for diabetic wound healing: myricetin and ceria nanozymes for enhanced reactive oxygen species management.

Regenerative biomaterials·2026
Same author

Prot-ΔΔG: Prediction of protein-protein binding affinity changes upon mutations with pre-training strategies.

Journal of computer-aided molecular design·2026
Same author

Association of baPWV and CD34<sup>+</sup> progenitor-derived exosomal hsa_circ_0093884 with Alzheimer's disease: mechanistic insights into the miR-375/SIX4 axis.

Frontiers in cell and developmental biology·2026
Same author

Type I Interferons Drive the Gastrointestinal Inflammatory Response in a Mouse Model of Parkinson's Disease.

Gastro hep advances·2026
Same author

Investigating driving behavior characteristics on freeway curves under accumulated sand environments: evidence from a driving simulator study.

Traffic injury prevention·2026
Same author

Bioactive Glass-Reinforced Composite Bone Adhesive with pH Modulation, Mechanical Strengthening, and Antibacterial and Osteogenic Functions.

ACS applied materials & interfaces·2026

相关实验视频

Updated: Jun 25, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

2.7K

实时驾驶风险预测使用基于自我注意的双向长期短期记忆网络,基于多源数据.

Zhuopeng Xie1, Yongfeng Ma2, Ziyu Zhang2

  • 1Jiangsu Key Laboratory of Urban ITS, School of Transportation, Southeast University, Nanjing 211189, China; School of Civil Engineering, Faculty of Engineering, University of Sydney, Darlington NSW 2008, Australia.

Accident; analysis and prevention
|May 26, 2024
PubMed
概括

本研究介绍了一种基于自我注意的双向长期短期记忆 (Att-Bi-LSTM) 模型,用于使用多源数据准确预测驾驶风险,显著优于现有方法.

关键词:
驱动风险 驱动风险是什么?驾驶模拟器驾驶模拟器长期短期记忆 长期短期记忆多个来源的数据数据.专注于自己的注意力

更多相关视频

Tactile Vibrating Toolkit and Driving Simulation Platform for Driving-Related Research
07:15

Tactile Vibrating Toolkit and Driving Simulation Platform for Driving-Related Research

Published on: December 18, 2020

4.5K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

520

相关实验视频

Last Updated: Jun 25, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

2.7K
Tactile Vibrating Toolkit and Driving Simulation Platform for Driving-Related Research
07:15

Tactile Vibrating Toolkit and Driving Simulation Platform for Driving-Related Research

Published on: December 18, 2020

4.5K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

520

科学领域:

  • 人工智能的人工智能
  • 运输安全运输安全
  • 机器学习 机器学习

背景情况:

  • 预测驾驶风险对于防止碰撞至关重要.
  • 现有的方法通常依赖于单个数据源,缺乏先进的建模或时间窗口分析.

研究的目的:

  • 提出一个新的Att-Bi-LSTM模型,用于多源驾驶风险预测.
  • 评估模型的性能与已建立的机器学习算法对比.

主要方法:

  • 从模拟测试中收集的多来源驾驶员数据 (人口统计,操作,视觉,生理,动力学).
  • 开发了一个Att-Bi-LSTM网络,并将其与CNN,CNN-LSTM,CatBoost,LightGBM和XGBoost进行了比较.
  • 利用观察,间隔和预测时间窗口用于模型输入/输出生成.

主要成果:

  • Att-Bi-LSTM模型实现了0.914的宏观平均F1得分,超过了所有比较模型.
  • 废除研究证实了Bi-LSTM层和自我注意机制的有效性.
  • 与仅使用动态数据相比,多源数据将F1得分提高了0.061.

结论:

  • 拟议的Att-Bi-LSTM模型提供了一种有效的驾驶风险预测方法.
  • 模型性能对时间窗口配置敏感,最佳的观察窗口提高了准确性.
  • 这项研究支持开发增强的先进驾驶辅助系统.