Fusing static driver traits with dynamic trajectories: a novel lane-changing intention prediction model
Measurement Science and Technology · 27.03.2026 · 37 (12) · 126303
DOI: 10.1088/1361-6501/ae5337 ↗Kurz gefasst
Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 27.03.2026. Verfasst von Weiqi Zhou und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 34 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Key Research and Development Program of Jiangsu Province.
Abstract (Originalsprache)
Accurately predicting the behavioral intentions of surrounding vehicles is essential for ensuring driving safety in autonomous driving systems. Existing methods often struggle to capture long-term features due to short input sequences, while directly extending sequence length can lead to a substantial increase in computational cost. To address these challenges, this paper proposes a lane-changing intention prediction model that integrates static driving characteristics. The model identifies driving styles through motion feature clustering and incorporates vehicle type as a static attribute to compensate for the lack of persistent vehicle-specific information characteristics in time-series data. Additionally, a bidirectional long short-term memory (Bi-LSTM) network is employed to extract features from the historical trajectories of the target vehicle, adjacent vehicles, and surrounding lane geometry. Finally, an attention mechanism is used to fuse static and dynamic features. Experiments conducted on the HighD dataset show that the proposed model improves prediction accuracy by 3.5%, 0.8%, and 1.23% for left lane-changing, lane-keeping, and right lane-changing tasks, respectively, achieving an overall accuracy of 99.37%, outperforming existing state-of-the-art methods.
Förderung
- National Natural Science Foundation of China
- Key Research and Development Program of Jiangsu Province
Zitieren
APA
Zhou, W., Bi, C., Li, Y., & Sun, K. (2026). Fusing static driver traits with dynamic trajectories: a novel lane-changing intention prediction model. Measurement Science and Technology, 37(12), 126303. https://doi.org/10.1088/1361-6501/ae5337
Vancouver
Zhou W, Bi C, Li Y, Sun K. Fusing static driver traits with dynamic trajectories: a novel lane-changing intention prediction model. Measurement Science and Technology. 2026;37(12):126303. doi:10.1088/1361-6501/ae5337
BibTeX
@article{Zhou2026,
author = {Zhou, Weiqi and Bi, Chenglong and Li, Yicheng and Sun, Kecheng},
title = {Fusing static driver traits with dynamic trajectories: a novel lane-changing intention prediction model},
journal = {Measurement Science and Technology},
year = {2026},
volume = {37},
number = {12},
pages = {126303},
doi = {10.1088/1361-6501/ae5337}
}
EUVIMED