Research on Physics-Informed Transformer for Space Object Shape Classification
Universe · 15.07.2026 · 12 (7) · 211
DOI: 10.3390/universe12070211 ↗Kurz gefasst
Fachartikel in der Zeitschrift Universe, online erschienen am 15.07.2026. Verfasst von Mengci Li und 10 weiteren Autor:innen. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key Laboratory of Space Security, National Natural Science Foundation of China, Key Laboratory of Intelligent Space TTC&O (Space Engineering University), Ministry of Education.
Abstract (Originalsprache)
Estimating the shape of space objects helps infer key characteristics such as object type, mass, and potential operational status, thereby providing critical decision-making support for space threat assessment. This paper proposes a Physics-Informed Transformer deep learning model for space object shape classification based on photometric time series data. The model achieves a classification accuracy of 89.22% on a hybrid dataset combining simulated and measured data, outperforming eight classical models with an average accuracy improvement of 3.61%. Ablation experiments demonstrate that the introduction of physical gating yields an average accuracy improvement of 1.43%. Using evaluation metrics including the confusion matrix, PR curve, ROC curve, and confidence distribution histogram, we demonstrate that the model possesses high accuracy, strong robustness, and good interpretability.
Förderung
- National Key Laboratory of Space Security
- National Natural Science Foundation of China
- Key Laboratory of Intelligent Space TTC&O (Space Engineering University), Ministry of Education
- Jilin Province Science and Technology Development Plan Item
- Chinese Academy of Sciences President’s International Fellowship Initiative
- Chinese Academy of Sciences National Astronomical Observatories Changchun Observatory 2025 Youth Exploration Fund
Zitieren
APA
Li, M., Zhang, L., Li, R., Li, Y., Deng, S., Guo, H., Zheng, H., Zhang, H., Zhao, Y., Deng, Z., & Zhu, R. (2026). Research on Physics-Informed Transformer for Space Object Shape Classification. Universe, 12(7), 211. https://doi.org/10.3390/universe12070211
Vancouver
Li M, Zhang L, Li R, Li Y, Deng S, Guo H, et al. Research on Physics-Informed Transformer for Space Object Shape Classification. Universe. 2026;12(7):211. doi:10.3390/universe12070211
BibTeX
@article{Li2026,
author = {Li, Mengci and Zhang, Laixian and Li, Rong and Li, Yingchun and Deng, Shiyu and Guo, Huichao and Zheng, Haijing and Zhang, Huaili and Zhao, Yang and Deng, Zhen and Zhu, Rui},
title = {Research on Physics-Informed Transformer for Space Object Shape Classification},
journal = {Universe},
year = {2026},
volume = {12},
number = {7},
pages = {211},
doi = {10.3390/universe12070211}
}
EUVIMED