A Deep Learning-Based Integrated Framework for Fault Diagnosis and Beam Recovery of Phased Array Antennas
IEEE Transactions on Antennas and Propagation · 2026-08 · 74 (8) · 7670-7682
DOI: 10.1109/tap.2026.3690196 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Antennas and Propagation, 2026-08. Verfasst von Qing Cai und 5 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Shandong Province Key Laboratory for Electromagnetic Control and Multifunctional Integration Technology of Aerospace Electromagnetic Functional Structure, Research Institute for Special Structures of Aeronautical Composites, Aviation Industry Corporation of China.
Abstract (Originalsprache)
A deep learning-based integrated framework is proposed for fault diagnosis and beam recovery of phased array antennas to ensure their beamforming and beam scanning performance. The proposed method comprises two core components: an amplitude–phase fault diagnosis network and a beam recovery network. The amplitude–phase fault diagnosis network takes the beam pattern as input and reconstructs the amplitude and phase excitations of faulty phased array antennas. By performing analyses on these reconstructed excitations, faulty excitations can be identified, thereby determining the number, location, and type of faulty elements in phased array antennas. Subsequently, the beam recovery network generates amplitude and phase excitations for beam recovery of phased array antennas with faulty elements, based on the current fault condition and desired beam direction. Experimental results on the $3 \times 3$ array demonstrate that the proposed amplitude–phase fault diagnosis network achieves an overall diagnostic accuracy of 99.76% under ideal conditions, with an average response time of 4.9 ms. For beam recovery, the proposed beam recovery network attains an average beam-pointing error (BPE) of 0.06° and an average response time of approximately 3.31 ms, and validation on the $10 \times 10$ array further reveals its considerable scalability and promising potential for engineering applications.
Förderung
- National Natural Science Foundation of China
- Shandong Province Key Laboratory for Electromagnetic Control and Multifunctional Integration Technology of Aerospace Electromagnetic Functional Structure
- Research Institute for Special Structures of Aeronautical Composites, Aviation Industry Corporation of China
Zitieren
APA
Cai, Q., Jia, Z. L., Yao, H. M., Wu, B., Liu, Y., & Zhang, H. H. (2026). A Deep Learning-Based Integrated Framework for Fault Diagnosis and Beam Recovery of Phased Array Antennas. IEEE Transactions on Antennas and Propagation, 74(8), 7670-7682. https://doi.org/10.1109/tap.2026.3690196
Vancouver
Cai Q, Jia ZL, Yao HM, Wu B, Liu Y, Zhang HH. A Deep Learning-Based Integrated Framework for Fault Diagnosis and Beam Recovery of Phased Array Antennas. IEEE Transactions on Antennas and Propagation. 2026;74(8):7670-7682. doi:10.1109/tap.2026.3690196
BibTeX
@article{Cai2026,
author = {Cai, Qing and Jia, Zheng Lang and Yao, He Ming and Wu, Bian and Liu, Ying and Zhang, Huan Huan},
title = {A Deep Learning-Based Integrated Framework for Fault Diagnosis and Beam Recovery of Phased Array Antennas},
journal = {IEEE Transactions on Antennas and Propagation},
year = {2026},
volume = {74},
number = {8},
pages = {7670--7682},
doi = {10.1109/tap.2026.3690196}
}
EUVIMED