杜龙龙,米洁,马超,赵泓樾

• 1:北京信息科技大学机电工程学院

摘要(Abstract)：

针对转速变化影响诊断精度、传统PC端方法存在时延与计算资源依赖的问题，提出了基于知识蒸馏与边缘计算的变转速行星齿轮箱故障诊断方法。首先，运用Welch变换提取振动信号特征。其次，以卷积神经网络-双向门控循环单元-注意力（convolutional neural network-bidirectional gated recurrent unit-attention, CNN-BiGRU-Attention）复合网络作为教师网络，以小规模一维CNN(one-dimensional CNN, 1D-CNN)作为学生模型，通过知识蒸馏，提升学生模型的诊断准确率。最后，将优化后的学生模型部署至树莓派平台，实现对行星齿轮箱的实时故障诊断。研究结果显示，所提方法能够有效提取振动信号特征，提高小型模型的准确率；同时，采用边缘计算替代传统PC端，实现了高效、准确且节约资源的故障诊断。

关键词(KeyWords)： 行星齿轮箱;故障诊断;知识蒸馏;边缘计算

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 杜龙龙,米洁,马超,赵泓樾

DOI: 10.16508/j.cnki.11-5866/n.2026.01.007

参考文献(References)：

• [1]贾会芳，张明.齿轮箱故障信号SFLA-DBN算法多传感器信号融合诊断[J/OL].机械设计与制造：1-6.(2025-09-29)[2025-10-20]. https：//doi. org/10. 19356/j. cnki. 1001-3997.20250928. 010.JIA H F,ZHANG M. Diagnosis of SFLA-DBN algorithm multisensor signal fusion for gearbox fault signals[J]. Machinery Design&Manufacture:1-6.(2025-09-29)[2025-10-20].https：//doi. org/10. 19356/j. cnki. 1001-3997. 20250928. 010.(in Chinese)
• [2]吴胜利，郑子润，邢文婷.基于多信息融合的INFO-VMDCNN的齿轮箱故障诊断方法[J].振动与冲击，2025,44(13):309-316.WU S L,ZHENG Z R,XING W T. Gearbox fault diagnosis method based on INFO-VMD-CNN of multi-information fusion[J]. Journal of Vibration and Shock,2025,44(13):309-316.(in Chinese)
• [3]YAO Y,ZHANG S,YANG S X,et al. Learning attention representation with a multi-scale CNN for gear fault diagnosis under different working conditions[J]. Sensors,2020,20(4):1233.
• [4]司兆峰，齐洪钢.知识蒸馏方法研究与应用综述[J].中国图象图形学报，2023,28(9):2817-2832.SI Z F,QI H G. Survey on knowledge distillation and its application[J]. Journal of Image and Graphics,2023,28(9):2817-2832.(in Chinese)
• [5]邵仁荣，刘宇昂，张伟，等.深度学习中知识蒸馏研究综述[J].计算机学报，2022,45(8):1638-1673.SHAO R R,LIU Y A,ZHANG W,et al. A survey of knowledge distillation in deep learning[J]. Chinese Journal of Computers,2022,45(8):1638-1673.(in Chinese)
• [6]李春林.面向旋转机械故障诊断的特征表达与智能识别方法研究[D].广州：广东工业大学，2025.LI C L. Feature representation and intelligent recognition methods for rotating machinery fault diagnosis[D]. Guangzhou:Guangdong University of Technology,2025.(in Chinese)
• [7]马超，郑鑫辉，王少红，等.基于SPWVD与知识蒸馏的行星变速器故障诊断研究[J].电子测量与仪器学报，2024,38(5):29-37.MA C,ZHENG X H,WANG S H,et al. Research on fault diagnosis of planetary transmission based on SPWVD and knowledge distillation[J]. Journal of Electronic Measurement and Instrumentation,2024,38(5):29-37.(in Chinese)
• [8]李屹，魏建国，江浩然.神经网络知识蒸馏方法综述[J].信息系统工程，2022(10):128-131.LI Y,WEI J G,JIANG H R. Overview of knowledge distillation methods for neural networks[J]. China CIO News,2022(10):128-131.(in Chinese)
• [9]ZHAO Y,YIN Y,GUI G. Lightweight deep learning based intelligent edge surveillance techniques[J]. IEEE Transactions on Cognitive Communications and Networking,2020,6(4):1146-1154.
• [10]吕欣然.基于深度神经网络的滚动轴承故障诊断方法研究[D].大连：大连交通大学，2024.L??X R. Research on fault diagnosis method of rolling bearing based on deep neural network[D]. Dalian:Dalian Jiaotong University,2024.(in Chinese)
• [11]JIN Z H,HAN Q C,ZHANG K,et al. An intelligent fault diagnosis method of rolling bearings based on Welch power spectrum transformation with radial basis function neural network[J]. Journal of Vibration and Control,2020,26(9/10):629-642.
• [12]LECUN Y,BOTTOU L,BENGIO Y,et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE,1998,86(11):2278-2324.
• [13]李练兵，高国强，陈伟光，等.考虑特征重组和BiGRUAttention-XGBoost模型的超短期负荷功率预测[J].现代电力，2025,42(3):571-581.LI L B,GAO G Q,CHEN W G,et al. Ultra short term load power prediction considering feature recombination and BiGRUAttention-XGBoost model[J]. Modern Electric Power,2025,42(3):571-581.(in Chinese)
• [14]邱宝林，马超，王少红.频响函数驱动的BYOL在行星变速器故障诊断中的应用[J].北京信息科技大学学报（自然科学版），2025,40(4):58-67.QIU B L,MA C,WANG S H. Application of BYOL driven by frequency response function in planetary gearbox fault diagnosis[J]. Journal of Beijing Information Science and Technology University(Science and Technology Edition),2025,40(4):58-67.(in Chinese)
• [15]毛荣珍，米洁，甄真，等.基于数据增强的多输出分类旋转机械复合故障诊断[J].北京信息科技大学学报（自然科学版），2025,40(1):94-102.MAO R Z,MI J,ZHEN Z,et al. Multi-output classification based on data augmentation for composite fault diagnosis of rotating machinery[J]. Journal of Beijing Information Science&Technology University(Science and Technology Edition),2025,40(1):94-102.(in Chinese)