基于RBVS和CBCNN的风机叶片故障检测和分类方法

doi:10.13229/j.cnki.jdxbgxb.20240617

吉林大学学报(工学版) ›› 2025, Vol. 55 ›› Issue (10): 3119-3130.doi: 10.13229/j.cnki.jdxbgxb.20240617

• 车辆工程·机械工程 • 上一篇

基于RBVS和CBCNN的风机叶片故障检测和分类方法

周求湛¹(),牟岩¹,武慧南¹,陈霄¹,汪锋²,李琛²,张雯²,刘萍萍³,王聪¹()

^1.吉林大学通信工程学院，长春 130012
^2.华锐风电科技（江苏）有限公司，江苏盐城 224056
^3.吉林大学计算机科学与技术学院，长春 130012

收稿日期:2024-06-03 出版日期:2025-10-01 发布日期:2026-02-03
通讯作者: 王聪 E-mail:13504465154@163.com;wangcong2020@jlu.edu.cn
作者简介:周求湛（1974-），男，教授，博士. 研究方向：微弱信号检测.E-mail：13504465154@163.com
基金资助:
盐城市重点研发计划（工业）项目(BE2023008);国家自然科学基金项目(62071199)

Fault detection and classification of wind turbine blades based on machine learning

Qiu-zhan ZHOU¹(),Yan MU¹,Hui-nan WU¹,Xiao CHEN¹,Feng WANG²,Chen LI²,Wen ZHANG²,Ping-ping LIU³,Cong WANG¹()

^1.College of Communication Engineering，Jilin University，Changchun 130012，China
^2.Sinovel Wind Power Technology（jiangsu） Co. ，Ltd. ，Yancheng 224056，China
^3.College of Computer Science and Technology，Jilin University，Changchun 130012，China

Received:2024-06-03 Online:2025-10-01 Published:2026-02-03
Contact: Cong WANG E-mail:13504465154@163.com;wangcong2020@jlu.edu.cn

摘要/Abstract

摘要：

为提高风机叶片故障检测时故障分类精度，提出了一种基于机器学习的风机叶片故障检测和分类方法。首先，将岭回归与蜂群优化算法（BSO）相结合提出了R-BSO特征选择算法，该算法用于筛选出最优特征子集。然后，将由R-BSO算法提取出的最佳特征组合输入基于Stacking策略的分类模型中得出分类结果，完成叶片故障检测RBVS算法的构建。最后，提出了一种基于卷积注意力机制（CBAM）的卷积神经网络（CNN）叶片故障分类算法CBCNN。实验结果表明：本文算法在风机叶片故障检测和分类上具有较好的性能。

关键词: 特征选择, 机器学习, Stacking, 卷积神经网络, 卷积注意力机制

Abstract:

To improve the classification accuracy of fault detection for wind-turbine blades， a machine-learning-based fault detection and classification method is proposed. First， a Ridge-regression-enhanced Brain Storm Optimization （R-BSO） feature-selection algorithm is developed to identify an optimal feature subset. The best feature combination extracted by R-BSO is then fed into a Stacking-based classifier to produce the final prediction， completing the RBVS blade-fault detection framework. Finally， a convolutional neural network equipped with a Convolutional Block Attention Module （CBAM）， termed CBCNN， is introduced for blade-fault classification. Experimental results demonstrate that the proposed algorithms achieve superior performance in both detection and classification of wind-turbine blade faults..

Key words: feature selection, machine learning, stacking, convolutional neural network, convolutional block attention module（CBAM）

中图分类号:

TP274

周求湛,牟岩,武慧南,陈霄,汪锋,李琛,张雯,刘萍萍,王聪. 基于RBVS和CBCNN的风机叶片故障检测和分类方法[J]. 吉林大学学报(工学版), 2025, 55(10): 3119-3130.

Qiu-zhan ZHOU,Yan MU,Hui-nan WU,Xiao CHEN,Feng WANG,Chen LI,Wen ZHANG,Ping-ping LIU,Cong WANG. Fault detection and classification of wind turbine blades based on machine learning[J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(10): 3119-3130.

图/表 16

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

表1

图11

图12

表2

图13

表3

参考文献 29

[1]	Anindya G, Sundaresan Mannur J, Schulz Mark J, et al. Structural health monitoring techniques for wind turbine blades[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2000, 85(3): 309-324.
[2]	Mayes R L. An experimental algorithm for detecting damage applied to the I-40 bridge over the Rio-grande [C]∥Smart Systems for Bridges, Structures, and Highways Conference on Smart Structures and Materials, Nashville, Tennessee, 1995: 219-225.
[3]	Dervilis N, Choi M, Taylor S G, et al. On damage diagnosis for a wind turbine blade using pattern recognition[J]. Journal of Sound and Vibration, 2014, 333(6): 1833-1850.
[4]	Sohn H, Worden K, Farrar C R. Novelty detection under changing environmental conditions[C]∥Smart Systems for Bridges, Structures, and Highways Conference on Smart Structures and Materials, Newport Beach, USA, 2001: 108-118.
[5]	Scholz M, Vigario R. Nonlinear PCA: a new hierarchical approach[C]∥The 10th European Symposium on Artificial Neural Networks, Bruges, Belgium, 2002:439-444.
[6]	Wang M H, Lu S D, Hsieh C C, et al. Fault detection of wind turbine blades using multi-channel CNN[J]. Sustainability (Switzerland), 2022, 14(3): No.1781.
[7]	宋柬霏. 小样本(医疗)数据的群智能特征选择算法研究[D]. 长春: 吉林大学通信工程学院, 2023.
	Song Jian-fei. Research on swarm intelligence feature selection algorithm for small sample (medical) data[D]. Changchun: College of Communication Engineering, Jilin University, 2023.
[8]	Jia H M, Li J D, Song W L, et al. Spotted hyena optimization algorithm with simulated annealing for feature selection[J]. IEEE Access, 2019, 7: 71943-71962.
[9]	Ou Y W, Tatsis K E. Vibration-based monitoring of a small-scale wind turbine blade under varying climate conditions. part Ⅰ: an experimental benchmark[J]. Structural Control & Health Monitoring, 2021,28(6): No.2660.
[10]	Philipp G, Lukas L. Non-uniqueness theory in sampled STFT phase retrieval[J]. SIAM Journal on Mathematical Analysis, 2023, 55(5): 4695-4726.
[11]	Gai J X, Zhang L H, Wei Z H. Spectrum sensing based on STFT-ImpResNet for cognitive radio[J]. Electronics, 2022, 11(15): No.2437.
[12]	姚欣宜. 地面运动目标振动感知与识别方法研究[D]. 长春: 吉林大学通信工程学院, 2022.
	Yao Xin-yi. Research on moving ground-target perception and recognition method based on vibration[D]. Changchun: College of Communication Engineering, Jilin University, 2022.
[13]	李婷, 付德义, 薛扬. 基于AE与STFT的变桨轴承裂纹诊断研究[J]. 振动、测试与诊断, 2021, 41(2): 299-303.
	Li Ting, Fu De-yi, Xue Yang. Research on crack diagnosis of pitch bearing based on AE and STFT[J]. Journal of Vibration, Measurement & Diagnosis, 2021, 42(2): 299-303.
[14]	Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[C]∥ Advances in Neural Information Processing Systems, Lake Tahoe, USA, 2012: 1097-1105.
[15]	Hubel D H, Wiesel T N. Receptive fields, binocular interaction and functional architecture in the cat´s visual cortex[J]. The Journal of Physiology, 1962, 160(1): 106-154.
[16]	Zhou Z H, Wu J, Tang W. Ensembling neural networks: many could be better than all[J]. Artificial Intelligence, 2002, 137(1-2): 239-263.
[17]	Vinay M B, Rekha K S. A model of saliency-based visual attention for rapid scene analysis[J]. International Journal of Recent Technology and Engineering, 2019, 7(6): 412-415.
[18]	冀泽宇. 分布式轻量化非侵入式电力负荷监测方法研究[D]. 长春: 吉林大学通信工程学院, 2023.
	Ji Ze-yu. Research on distributed lightweight non-intrusive power load monitoring method[D]. Changchun: College of Communication Engineering, Jilin University, 2023.
[19]	高广尚. 深度学习推荐模型中的注意力机制研究综述[J]. 计算机工程与应用, 2022, 58(9): 9-18.
	Gao Guang-shang. Survey on attention mechanisms in deep learning recommendation models[J]. Computer Engineering and Applications, 2022, 58(9): 9-18.
[20]	Hu J, Shen L, Albanie S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023.
[21]	Sanghyun W, Jongchan P, Joon-young L, et al. CBAM: convolutional block attention module[J].Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2018, 11211: 3-19.
[22]	王涛. 基于随机森林的机载设备异常振动故障检测方法[J]. 电子设计工程, 2023, 31(8): 119-122, 127.
	Wang Tao. Abnormal vibration fault detection method of airborne equipment based on random forest[J]. Electronic Design Engineering,2023,31(8): 119-122, 127.
[23]	马康原, 李忠虎, 王金明, 等. 基于XGBoost模型的风电机组故障检测[J].内蒙古科技大学学报, 2023(4): 349-353.
	Ma Kang-yuan, Li Zhong-hu, Wang Jin-ming, et al. Wind turbine fault detection based on XGBoost model[J]. Journal of Inner Mongolia University of Science and Technology,2023(4): 349-353.
[24]	陈宇韬, 唐明珠, 吴华伟, 等.基于极端随机森林的大型风电机组发电机故障检测[J].湖南电力, 2019, 39(6): 45-51.
	Chen Yu-tao, Tang Ming-zhu, Wu Hua-wei, et al. Fault detection based on extreme random forest for large wind turbine generators[J]. Hunan Electric Power, 2019, 39(6): 45-51.
[25]	梁川. 基于特征提取和随机森林的风机故障诊断[J].科学技术创新, 2020(26): 55-58.
	Liang Chuan. Wind turbine fault diagnosis based on feature extraction and random forest[J]. Scientific and Technological Innovation, 2020(26): 55-58.
[26]	张好雨. 基于SCADA数据的风机叶片覆冰故障检测与预测[D]. 北京: 北京交通大学机械工程学院,2022.
	Zhang Hao-yu. Detection and prediction of wind turbine blade icing fault based on SCADA data[D]. Beijing: College of Mechanical Engineering, Beijing Jiaotong University, 2022.
[27]	杨亚男. 基于SCADA数据的风机叶片故障诊断系统研究[D]. 北京: 华北电力大学控制与计算机工程学院, 2020.
	Yang Ya-nan. Research on fan blade fault diagnosis system based on SCADA data[D]. Beijing: College of Control and Computer Engineering, North China Electric Power University, 2020.
[28]	刘天奇, 贾明峰, 赵薇. 基于振动信号的风机叶片故障检测[J]. 电力设备管理, 2023(20): 90-92.
	Liu Tian-qi, Jia Ming-feng, Zhao Wei. Vibration signal-based fault detection of wind turbine blades[J]. Power equipment management, 2023(20): 90-92.
[29]	于嵩浩. 风力发电机组叶片健康智能声学监测系统[D]. 北京: 北京邮电大学自动化学院, 2020.
	Yu Song-hao. Intelligent acoustic monitoring system for wind turbine blade health[D]. Beijing: School of Automation, Beijing University of Posts and Telecommunications, 2020.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

特征算法	Accuracy	Precision	Recall	F₁-score
R-BSO	0.893 7	0.894 4	0.893 5	0.899 8
RFE	0.843 7	0.826 4	0.872 6	0.848 9
LASSO	0.851 5	0.847 0	0.860 2	0.853 6
Ridge	0.884 3	0.887 5	0.881 9	0.884 7
BSO	0.887 5	0.876 5	0.893 2	0.889 9
ReliefF	0.867 1	0.855 8	0.885 0	0.870 2

评价指标	RBVS	随机森林	XGBoost	极度随机树	ReliefF RF	RFXGB	LGBXGB
Accuracy	0.938 4	0.909 8	0.869 6	0.924 1	0.915 2	0.908 0	0.912 5
Precision	0.932 9	0.923 1	0.879 1	0.922 3	0.918 1	0.907 1	0.907 9
Recall	0.950 4	0.902 6	0.870 1	0.933 3	0.919 7	0.917 9	0.926 5
F₁-score	0.941 6	0.912 7	0.874 6	0.927 8	0.918 9	0.912 5	0.917 1
AUROC	0.981 4	0.967 9	0.951 9	0.968 4	0.971 5	0.970 4	0.971 2
AUPRC	0.979 1	0.965 4	0.951 6	0.962 9	0.970 3	0.970 8	0.971 9

	CBCNN	S-CNN	LSTM	CNN
验证集	0.990 0	0.880 6	0.689 6	0.596 5
测试集	0.901 3	0.816 5	0.668 7	0.536 5

基于RBVS和CBCNN的风机叶片故障检测和分类方法

Fault detection and classification of wind turbine blades based on machine learning

RICH HTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

图/表 16

参考文献 29

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	侯越,张鑫,武月. 基于时空动态约束图反馈的交通流预测[J]. 吉林大学学报(工学版), 2026, 56(1): 183-198.
[2]	田婧,马社强,宋现敏,赵丹,陈发城. 稀疏数据下的交通状态估计自适应卷积网络[J]. 吉林大学学报(工学版), 2025, 55(8): 2579-2587.
[3]	肖红,刘显德. 基于混合智能的学习状态实时采集与动态分析方法[J]. 吉林大学学报(工学版), 2025, 55(7): 2402-2408.
[4]	冯志刚,王首起,于明月. 基于变分模态提取及轻量级网络的滚动轴承故障诊断[J]. 吉林大学学报(工学版), 2025, 55(6): 1883-1891.
[5]	周丰丰,郭喆,范雨思. 面向不平衡多组学癌症数据的特征表征算法[J]. 吉林大学学报(工学版), 2025, 55(6): 2089-2096.
[6]	杜睿山,王紫珊. 基于时空注意力的多视角人脸表情识别算法[J]. 吉林大学学报(工学版), 2025, 55(6): 2097-2102.
[7]	梅生启,刘晓东,王兴举,李旭峰,武腾,程相旭. 基于参数相关性分析和机器学习算法的高强混凝土徐变预测[J]. 吉林大学学报(工学版), 2025, 55(5): 1595-1603.
[8]	赵秀芝,谢德红. 基于噪声鲁棒性特征提取的普洱茶品种鲁棒判别方法[J]. 吉林大学学报(工学版), 2025, 55(5): 1756-1762.
[9]	金庆良,周鑫森,陈翼,吴承文. 基于群智能增强核极限学习机的创新人才预测模型[J]. 吉林大学学报(工学版), 2025, 55(5): 1763-1771.
[10]	孟祥海,王国锐,张明扬,田毕江. 基于选择集成的山区高速事故预测模型[J]. 吉林大学学报(工学版), 2025, 55(4): 1298-1306.
[11]	戴银飞,周秀贞,刘玉宝,刘志远. 基于CAN总线数据的车载网络入侵检测系统[J]. 吉林大学学报(工学版), 2025, 55(3): 857-865.
[12]	罗维薇,刘长龙,雷琴. 基于预处理层增强和注意力机制的空域图像隐写分析[J]. 吉林大学学报(工学版), 2025, 55(12): 4024-4033.
[13]	胡宏宇,张争光,曲优,蔡沐雨,高菲,高镇海. 基于双分支和可变形卷积网络的驾驶员行为识别方法[J]. 吉林大学学报(工学版), 2025, 55(1): 93-104.
[14]	汪豪,赵彬,刘国华. 基于时间和运动增强的视频动作识别[J]. 吉林大学学报(工学版), 2025, 55(1): 339-346.
[15]	张磊,焦晶,李勃昕,周延杰. 融合机器学习和深度学习的大容量半结构化数据抽取算法[J]. 吉林大学学报(工学版), 2024, 54(9): 2631-2637.