李亚超,刘昊煜,徐皓康,官宇晗,祁湛桐,顾煜炯

• 1:国家能源集团花园电厂
• 2:华北电力大学能源动力与机械工程学院

摘要(Abstract)：

滑动轴承的早期故障具有隐秘性，为了准确预测其振动幅值，提出了一种结合YOLOv8优化的CBAM(convolutional block attention module)的深度学习模型，在Backbone和Neck之间嵌入CBAM模块以提升模型对重要振动特征的关注度，同时采用改进的完全交并比损失函数，提高目标检测精度。同时，考虑到振动数据的非线性、非稳态特性，在模型中添加经验模态分解（empirical mode decomposition,EMD）方法对振动状态数据进行预测，以提高预测的准确性。结果表明：该方法在600MW汽轮机运行数据集上相较于传统YOLOv8和YOLOv7，在目标检测准确率上分别提升2.85百分点和8.50百分点，均方根误差和平均绝对误差均有所减少；此外，在高噪声环境下，该模型的误差波动较传统方法降低30%，表现出更强的泛化能力和稳定性。

关键词(KeyWords)： 注意力机制;汽轮机振动;YOLO;经验模态分解

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 李亚超,刘昊煜,徐皓康,官宇晗,祁湛桐,顾煜炯

DOI: 10.19666/j.rlfd.202412263

参考文献(References)：

• [1]李晓博,舒进,牛瑞杰,等.基于主动电磁控制的滑动轴承-转子系统自激振动抑制数值仿真[J].热力发电,2020, 49(9):80-86.LI Xiaobo, SHU Jin, NIU Ruijie, et al. Numerical simulation of self-excited vibration suppression of oil-film bearing-rotor system based on active electro-magnetic control[J]. Thermal Power Generation,2020, 49(9):80-86.
• [2] DAI J, TIAN L, CHANG H. An intelligent diagnostic method for wear depth of sliding bearings based on MGCNN[J]. Machines, 2024, 12(4):266.
• [3] LI Q, ZHANG W, CHEN F, et al. Fault diagnosis of nuclear power plant sliding bearing-rotor systems using deep convolutional generative adversarial networks[J].Nuclear Engineering and Technology, 2024, 56(8):2958-2973.
• [4] CHEN F, WANG X, ZHU Y, et al. Time-frequency transformer with shifted windows for journal bearing-rotor systems fault diagnosis under multiple working conditions[J]. Measurement Science and Technology, 2023, 34(8):085121.
• [5] TANG H, LIAO Z, OZAKI Y, et al. Stepwise intelligent diagnosis method for rotor system with sliding bearing based on statistical filter and stacked auto-encoder[J].Applied Sciences, 2020, 10(7):2477.
• [6] SONG L, WANG H, CHEN P. Step-by-step fuzzy diagnosis method for equipment based on symptom extraction and trivalent logic fuzzy diagnosis theory[J].Ieee Transactions on Fuzzy Systems, 2018, 26(6):3467-3478.
• [7] HUANG T, FU S, FENG H, et al. Bearing fault diagnosis based on shallow multi-scale convolutional neural network with attention[J]. Energies, 2019, 12(20):3937.
• [8] LUO J, ZHANG X. Convolutional neural network based on attention mechanism and Bi-LSTM for bearing remaining life prediction[J]. Applied Intelligence, 2022,52(1):1076-1091.
• [9]吴静然,丁恩杰,崔冉,等.采用多尺度注意力机制的旋转机械故障诊断方法[J].西安交通大学学报, 2020,54(2):51-58.WU Jingran, DING Enjie, CUI Ran, et al. A diagnostic approach for rotating machinery using multi-scale feature attention mechanism[J]. Journal of Xi’an Jiaotong University, 2020, 54(2):51-58.
• [10]武福,蒋鹏民,李忠学,等.基于轻量化YOLOv8s的轨道扣件状态检测方法[J].北京交通大学学报, 2024,48(5):59-68.WU Fu, JIANG Pengmin, LI Zhongxue, et al.Light-weight detection method for track fastener status based on improved YOLOv8s[J]. Journal of Beijing Jiaotong University, 2024, 48(5):59-68.
• [11] DING P, ZHAN H, YU J, et al. A bearing surface defect detection method based on multi-attention mechanism Yolov8[J]. Measurement Science and Technology, 2024,35(8):086003.
• [12] WANG H, SHEN Q, DAI Q, et al. Evolutionary variational YOLOv8 network for fault detection in wind turbines[J]. Computers Materials&Continua, 2024,80(1):625-642.
• [13] LIU B, ZHAO Y, CHEN B, et al. CAC-YOLOv8:real-time bearing defect detection based on channel attenuation and expanded receptive field strategy[J].Measurement Science and Technology, 2024, 35(9):096004.
• [14] LIU M, ZHANG M, CHEN X, et al. YOLOv8-LMG:An improved bearing defect detection algorithm based on YOLOv8[J]. Processes, 2024, 12(5):930.
• [15]王普,李天垚,高学金,等.分层自适应小波阈值轴承故障信号降噪方法[J].振动工程学报, 2019, 32(3):548-556.WANG Pu, LI Tianyao, GAO Xuejin, et al. Bearing fault signal denoising method of hierarchical adaptivewavelet threshold function[J]. Journal of Vibration Engineering,2019, 32(3):548-556.
• [16] BAKIR T, BOUSSAID B, HAMDAOUI R, et al. Fault detection in wind turbine system using wavelet transform:Multi-resolution analysis[C]. 2015 IEEE 12th International Multi-Conference on Systems, Signals&Devices(SSD15), Mahdia, Tunisia, 16-19 March 2015.
• [17] HUANG Y, WANG K, DENG Z, et al. Synchronous averaging with sliding narrowband filtering for low-speed bearing fault diagnosis[J]. Journal of Sound and Vibration, 2024, 586:118503.
• [18] XIN Y, LI S. Novel data-driven short-frequency mutual information entropy threshold filtering and its application to bearing fault diagnosis[J]. Measurement Science and Technology, 2019, 30(11):115006.
• [19] CHEN J, PAN J, LI Z, et al. Generator bearing fault diagnosis for wind turbine via empirical wavelet transform using measured vibration signals[J].Renewable Energy, 2016, 89:80-92.
• [20] ZHANG Q, WEI X, WANG Y, et al. Convolutional neural network with attention mechanism and visual vibration signal analysis for bearing fault diagnosis[J].Sensors, 2024, 24(6):1831.
• [21] QIN H, PAN J, LI J, et al. Fault diagnosis method of rolling bearing based on CBAM_ResNet and ACON activation function[J]. Applied Sciences-Basel, 2023,13(13):7593.
• [22] XU Z, WANG Y, XIONG W, et al. A novel attentional feature fusion with inception based on capsule network and application to the fault diagnosis of bearing with small data samples[J]. Machines, 2022, 10(9):789.
• [23]杨博,胡珍珍.基于YOLOv8n改进算法的自动驾驶目标检测[J/OL].控制工程[2025-02-05]. DOI:10.14107/j. cnki.kzgc.20240494.YANG Bo, HU Zhenzhen. Automatic driving target detection based on YOLOv8n improved algorithm[J].Control Engineering of China[2025-02-05]. DOI:10.14107/j.cnki. kzgc.20240494.
• [24]钟帅,王丽萍.改进YOLOv8的无人机航拍图像目标检测方法[J/OL].航空兵器, 1-11[2025-02-07].https://kns.cnki.net/kcms/detail/41.1228.TJ.20241218.1415.003.html.ZHONG Shuai, WANG Liping. Improved YOLOv8obiect detection method for drone aerial images[J/OL].Aero Weaponry, 1-11[2025-02-07]. https://kns.cnki.net/kcms/detail/41.1228.TJ.20241218.1415.003.html.
• [25]王璐,张爽.基于EMD和FFT的自适应X射线脉冲星信号降噪方法[J].电波科学学报, 2025, 40(2):381-394.WANG Lu, ZHANG Shuang. Adaptive denoising method of X-ray pulsar signal based on EMD and FFT[J]. Chinese Journal of Radio Science, 2025, 40(2):381-394.