中国舰船研究2026，Vol.21Issue(3)：263-271,9.DOI:10.19693/j.issn.1673-3185.04384
基于频带注意力网络的齿轮箱小样本故障诊断方法
Small sample gearbox fault diagnosis method based on a frequency band attention network
摘要
Abstract
[Objective]Deep learning-based fault diagnosis methods typically require large amounts of fault data.To enable accurate gearbox fault diagnosis in small-sample scenarios,a novel diagnosis method based on a frequency band attention network is proposed.[Method]First,a reconstruction-encoding layer is used to transform vibration signals into sub-band encoded signals that are more suitable for classification.Then,an in-trinsic band attention layer is designed to effectively extract salient time-frequency features from the sub-band encoded signals.Finally,a multi-feature fusion module is used to integrate the extracted time-frequency fea-tures for fault recognition in small-sample conditions.[Results]Experimental results on a gearbox fault simul-ation platform show that the proposed method achieves a fault diagnosis accuracy of 99.85%in small-sample conditions,surpassing existing benchmark models.[Conclusion]These findings can provide a valuable ref-erence for gearbox fault diagnosis in small-sample conditions.关键词
齿轮箱/旋转机械/故障分析/频带注意力/信号重构/小样本/深度学习Key words
gearboxes/rotating machinery/failure analysis/frequency band attention network/signal re-construction/small-sample/deep learning分类
交通工程引用本文复制引用
涂序源,邓琪,张祚庥,王子木直,吴军..基于频带注意力网络的齿轮箱小样本故障诊断方法[J].中国舰船研究,2026,21(3):263-271,9.基金项目
华中科技大学交叉研究支持计划项目(2024JCYJ028) （2024JCYJ028）
国家自然科学基金资助项目(523B2100) （523B2100）
