EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A framework for detecting and predicting highway traffic anomalies via multimodal fusion and heterogeneous graph neural networks

Shaowei Sun iD, Mingzhou Liu

PLOS One · 25.06.2025 · 20 (6) · e0326313

DOI: 10.1371/journal.pone.0326313 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 25.06.2025. Verfasst von Shaowei Sun und 1 weiteren Autor:innen. Beteiligte Einrichtung: School of Mechanical Engineering, Hefei University of Technology, Hefei, Anhui, China. Hauptthemen laut MEDLINE: Neural Networks, Computer. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: This work was supported by the Science and Technology Project of Housing and Urban-Rural Construction of Anhui Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents a novel framework for detecting and predicting abnormal traffic events on highways. Current traffic monitoring systems often rely on single data sources, which limits their detection accuracy and robustness in complex environments. To address these limitations, we propose a framework based on multimodal deep fusion and heterogeneous graph neural networks (HGNNs), incorporating an Ensemble Contrastive Pessimistic Likelihood Estimation (CPLE) algorithm to optimize performance. The framework integrates static and dynamic traffic data, such as video images, traffic flow, vehicle speed, and tunnel weather conditions. Experimental results demonstrate that the model performs well in various scenarios, showing significant improvement in accuracy and stability over existing models like AGC-LSTM and AttentionDeepST. For instance, the proposed MHGNN-CPLE model achieves superior accuracy and F1 score in static detection tasks while maintaining high accuracy under different noise levels in dynamic detection scenarios. This study provides a significant advancement in traffic event analysis by effectively combining multimodal data and leveraging HGNNs to capture complex spatiotemporal dependencies.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerAlgorithmsHumansGraph Neural Networks

Autor:innen und Institutionen

  • School of Mechanical Engineering, Hefei University of Technology, Hefei, Anhui, China.
  • Intelligent Transportation Engineering Institute, Anhui Transport Consulting & Design Institute Co., Ltd, Hefei, Anhui, China.

ORCID-Profile: Shaowei Sun

Förderung

  • This work was supported by the Science and Technology Project of Housing and Urban-Rural Construction of Anhui Province

Zitieren

APA

Sun, S., & Liu, M. (2025). A framework for detecting and predicting highway traffic anomalies via multimodal fusion and heterogeneous graph neural networks. PLOS One, 20(6), e0326313. https://doi.org/10.1371/journal.pone.0326313

Vancouver

Sun S, Liu M. A framework for detecting and predicting highway traffic anomalies via multimodal fusion and heterogeneous graph neural networks. PLOS One. 2025;20(6):e0326313. doi:10.1371/journal.pone.0326313

BibTeX
@article{Sun2025,
  author = {Sun, Shaowei and Liu, Mingzhou},
  title = {A framework for detecting and predicting highway traffic anomalies via multimodal fusion and heterogeneous graph neural networks},
  journal = {PLOS One},
  year = {2025},
  volume = {20},
  number = {6},
  pages = {e0326313},
  doi = {10.1371/journal.pone.0326313}
}
Kostenlos starten →