基于多模态时频图融合的风电机组齿轮箱故障诊断方法

陈阳; 刘永前; 韩爽; 王罗

doi:10.16513/j.2096-2185.DE.2308303

PDF(1747 KB)

分布式能源 ›› 2023, Vol. 8 ›› Issue (3) : 17-23. DOI: 10.16513/j.2096-2185.DE.2308303

学术研究

基于多模态时频图融合的风电机组齿轮箱故障诊断方法

陈阳 ¹ ,
刘永前 ¹ ,
韩爽 ¹ ,
王罗 ²

作者信息 +

Fault Diagnosis Method of Wind Turbine Gearbox Based on Multi-Mode Time-Frequency Image Fusion

Author information +

文章历史 +

摘要

特征提取是故障诊断问题的关键，风电机组齿轮箱负载多变，振动数据量大、信息密度低，导致多模态融合诊断方法的特征学习性能和计算成本难以兼顾。为此，提出了基于多模态时频图融合的风电机组齿轮箱故障诊断方法。首先，使用小波包分解算法对齿轮箱振动数据集进行故障特征分析；然后，将所得分解子信号和原始信号分别转化为二维时频图像，形成体现小波域和时频域的互补故障特征；最后，利用卷积神经网络(convolutional neural networks, CNN)分别学习图像纹理的深层特征并进行特征融合，训练基于CNN-ViT(vision transformer)的多模态故障诊断模型。以凯斯西储大学齿轮箱轴承数据进行验证，所提模型在变负载和未知故障下，相比其他单模态和多模态方法具有较高的准确率，时频图融合方法能够以较低计算成本实现小波域和时频域双模态的故障特征学习。

Abstract

Feature extraction is the key to the fault diagnosis problem, and the variable load of wind turbine gearboxes, large amount of vibration data and low information density make it difficult to balance the feature learning performance and computational cost of multimodal fusion diagnosis method. To this end, a wind turbine gearbox fault diagnosis method based on multimodal time-frequency map fusion is proposed. First, a wavelet packet decomposition algorithm is used to analyze the fault features of the gearbox vibration data set; then, the decomposed sub-signal and the original signal are transformed into two-dimensional time-frequency images to form complementary fault features reflecting the wavelet domain and the time-frequency domain; finally, a convolutional neural network (CNN) is used to learn the deep features of the image texture and fuse them separately to train a CNN-ViT (vision transformer) multimodal fault diagnosis model. The proposed model is validated with Case Western Reserve University gearbox bearing data, and has higher accuracy compared with other unimodal and multimodal methods under variable load and unknown faults, and the time-frequency map fusion method can realize the fault feature learning in both wavelet domain and time-frequency domain dual modality with lower computational cost.

导出引用

陈阳, 刘永前, 韩爽, 等. 基于多模态时频图融合的风电机组齿轮箱故障诊断方法[J]. 分布式能源. 2023, 8(3): 17-23 https://doi.org/10.16513/j.2096-2185.DE.2308303

Yang CHAN, Yongqian LIU, Shuang HAN, et al. Fault Diagnosis Method of Wind Turbine Gearbox Based on Multi-Mode Time-Frequency Image Fusion[J]. Distributed Energy Resources. 2023, 8(3): 17-23 https://doi.org/10.16513/j.2096-2185.DE.2308303

中图分类号： TK81

参考文献

列表( 原文顺序 | 文献年度倒序 | 文中引用次数倒序 ) 可视化分析

[1]	李辉，胡姚刚，李洋，等. 大功率并网风电机组状态监测与故障诊断研究综述[J]. 电力自动化设备，2016, 36(1): 6-16. LI Hui, HU Yaogang, LI Yang, et al. Research review on condition monitoring and fault diagnosis of high power grid-connected wind turbines[J]. Electric Power Automation Equipment, 2016, 36(1): 6-16. 本文引用 [1]

[2]	张真真，吴立东，陈晓敏，等. 基于支持向量机的风电机组变桨系统故障诊断[J]. 分布式能源，2021, 6(3): 70-75. ZHANG Zhenzhen, WU Lidong, CHEN Xiaomin, et al. Fault diagnosis of wind turbine pitch system based on support vector machine[J]. Distributed Energy, 2021, 6(3): 70-75. 本文引用 [1]

[3]	KIM K, PARTHASARATHY G, ULUYOL O, et al. Use of SCADA data for failure detection in wind turbines[C]//Presented at the 2011 Energy Sustainability Conference and Fuel Cell Conference, 7-10 August 2011, Washington, D. C. American Society of Mechanical Engineers Digital Collection, 2011. 本文引用 [1]

[4]	陈雪峰，郭艳婕，许才彬，等. 风电装备故障诊断与健康监测研究综述[J]. 中国机械工程，2020, 31(2): 175-189. CHEN Xuefeng, GUO Yanjie, XU Caibin, et al. Review of fault diagnosis and monitoring for wind power equipment[J]. China Mechanical Engineering, 2020, 31(2): 175-189. 本文引用 [1]

[5]	李浪，刘辉海，赵洪山. 风力发电机振动监测与故障诊断方法综述[J]. 电网与清洁能源，2017, 33(8): 94-100, 108. LI Lang, LIU Huihai, ZHAO Hongshan. Review of vibration monitoring and fault diagnosis methods of wind turbines[J]. Power Grid and Clean Energy, 2017, 33(8): 94-100, 108. 本文引用 [1]

[6]	PICHIKA S N, MEGANAA G, GEETHA R S, et al. Multi-component fault classification of a wind turbine gearbox using integrated condition monitoring and hybrid ensemble method approach[J]. Applied Acoustics, 2022, 195. 本文引用 [1]

[7]	LIANG P, DENG C, YUAN X, et al. A deep capsule neural network with data augmentation generative adversarial networks for single and simultaneous fault diagnosis of wind turbine gearbox[J]. ISA Transactions, 2022. 本文引用 [1]

[8]	李东东，刘宇航，赵阳，等. 基于改进生成对抗网络的风机行星齿轮箱故障诊断方法[J]. 中国电机工程学报，2021, 41(21): 7496-7507. LI Dongding, LIU Yuhang, ZHAO Yang, et al. Fault diagnosis method of wind turbine planetary gearbox based on improved generative adversarial network[J]. Proceedings of the CSEE, 2021, 41(21): 7496-7507. 本文引用 [1]

[9]	ZHANG K, TANG B, DENG L, et al. A hybrid attention improved ResNet based fault diagnosis method of wind turbines gearbox[J]. Measurement, 2021, 179: 109491. 本文引用 [1]

[10]	JIANG G, JIA C, NIE S, et al. Multiview enhanced fault diagnosis for wind turbine gearbox bearings with fusion of vibration and current signals[J]. Measurement, 2022, 196. 本文引用 [1]

[11]	TANG Z, WANG M, OUYANG T, et al. A wind turbine bearing fault diagnosis method based on fused depth features in time-frequency domain[J]. Energy Reports, 2022, 8. 本文引用 [1]

[12]	薛阳，雷文平，岳帅旭，等. 多模态学习方法在滚动轴承故障诊断中的应用[J]. 机械科学与技术，2022, 41(8): 1149-1153. XUE Yang, LEI Wenping, YUE Shuaixu, et al. Application of multimodal deep learning methods in rolling bearing fault diagnosis[J]. Mechanical Science and Technology for Aerospace Enginering, 2022, 41(8): 1149-1153. 本文引用 [1]

[13]	马临超，杨捷，肖鹏，等. 基于新型卷积神经网络的非侵入式负载监测方法[J]. 智慧电力，2022, 50(4): 96-102. MA Linchao, YANG Jie, XIAO Peng, et al. Non-invasive load monitoring method based on novel convolutional neural network[J]. Smart Power, 2022, 50(4): 96-102. 本文引用 [1]

[14]	蒋爱国，符培伦，谷明，等. 基于多模态堆叠自动编码器的感应电机故障诊断[J]. 电子测量与仪器学报，2018, 32(8): 17-23. JIANG Aiguo, FU Peilun, GU Ming, et al. Induction motor fault diagnosis based on multimodal stacked auto-encoder[J]. Journal of Electronic Measurement and Instrumentation, 2018, 32(8): 17-23. 本文引用 [1]

[15]	付晓敏. 基于小波变换的改进混合蛙跳－差分进化－神经网络预测模型的短期风速预测[J]. 分布式能源，2021, 6(6): 38-44. FU Xiaomin. Improved hybrid frog-hopping-differential evolutionary-neural network prediction model based on wavelet transform for short-term wind speed prediction[J]. Distributed Energy, 2021, 6(6): 38-44. 本文引用 [1]

[16]	张利慧，李晓波，俎海东，等. 风力发电机组滚动轴承故障振动诊断方法综述[J]. 内蒙古电力技术，2022, 40(3): 7-12. ZHANG Lihui, LI Xiaobo, ZU Haidong, et al. Overview of fault vibration diagnosis methods for rolling bearing of wind turbine units[J]. Inner Mongolia Electric Power, 2022, 40(3): 7-12. 本文引用 [1]

[17]	黄树帮，陈耀，金宇清. 碳中和背景下多通道特征组合超短期风电功率预测[J]. 发电技术，2021, 42(1): 60-68. HUANG Shubang, CHEN Yao, JIN Yuqing. A multi-channel feature combination model for ultra-short-term wind power prediction under carbon neutral background[J]. Power Generation Technology, 2021, 42(1): 60-68. 本文引用 [1]

[18]	高阳，谢丽蓉，叶家豪，等. 自适应提升及预测误差修正的风电功率超短期预测[J]. 智慧电力，2022, 50(8): 14-21. GAO Yang, XIE Lirong, YE Jiahao, et al. Ultra-short-term wind power prediction based on adaptive lifting and prediction error correction[J]. Smart Power, 2022, 50(8): 14-21. 本文引用 [1]

[19]	崔宝珍，王斌，任川，等. 应用时频图像纹理特征的行星齿轮故障诊断[J]. 振动.测试与诊断，2022, 42(6): 1141-1146, 1245. CUI Baozhen, WANG Bin, REN Chuan, et al. Fault diagnosis of planetary gear used time-frequency image texture features[J]. Journal of Vibration. Measurement and Diagnosis, 2022, 42(6): 1141-1146, 1245. 本文引用 [1]

[20]	JIAO J, ZHAO M, LIN J, et al. A comprehensive review on convolutional neural network in machine fault diagnosis[J]. Neurocomputing, 2020, 417. 本文引用 [1]

[21]	任泽裕，王振超，柯尊旺，等. 多模态数据融合综述[J]. 计算机工程与应用，2021, 57(18): 49-64. REN Zeyu, WANG Zhenchao, KE Zunwang, et al. Survey of multimodal data fusion[J]. Computer Engineering and Applications, 2021, 57(18): 49-64. 本文引用 [1]

[22]	李世玲，李治，李合生. 基于小波包能量特征的滚动轴承故障监测方法[J]. 系统仿真学报，2003, 15(1): 76-80, 83. LI Shiling, LI Zhi, LI Hesheng. The method of roller bearing fault monitoring based on wavelet packet energy feature[J]. Journal of System Simulation, 2003, 15(1): 76-80, 83. 本文引用 [2]

[23]	崔玲丽，安加林，王鑫，等. 基于变尺度Lempel-Ziv的滚动轴承损伤程度评估方法[J]. 振动工程学报，2022, 35(5): 1250-1258. CUI Lingli, AN Jialin, WANG Xin, et al. Evaluation method of rolling bearing damage degree based on variable scale lempel-Ziv[J]. Journal of Vibration Engineering, 2022, 35(5): 1250-1258. 本文引用 [2]