Spatial-FineDef：融合多尺度感知与自适应增强的风电机组叶片小缺陷检测方法

年嘉伟; 王梓懿; 唐灵犀; 房方

doi:10.12263/DZXB.20250999

您当前的位置：

首页 >

文章列表页 >

Spatial-FineDef：融合多尺度感知与自适应增强的风电机组叶片小缺陷检测方法

学术论文 | 更新时间：2026-05-28

- Spatial-FineDef：融合多尺度感知与自适应增强的风电机组叶片小缺陷检测方法
- Spatial-FineDef: An Approach for Detecting Small Defects in Wind Turbine Blades that Integrate Multi-Scale Perception and Adaptive Enhancement
- 电子学报 2026年54卷第2期页码：785-798
- 作者机构：
  
  1.新能源电力系统全国重点实验室（华北电力大学），北京 102206
  2.华北电力大学控制与计算机工程学院，北京 102206
- 作者简介：
  
  [ "年嘉伟男，1995年生，博士。主要研究方向为智能发电、故障检测。 E-mail: jwnian@ncepu.edu.cn" ]
  [ "王梓懿男，2002年生，硕士研究生。主要研究方向为风电机组智能检测。 E-mail: ziyiwang@ncepu.edu.cn" ]
  [ "唐灵犀女，2001年生，硕士研究生。主要研究方向为深度学习、故障诊断。 E-mail: tanglingxi_edu@163.com" ]
  [ "房方男，1976年生，博士，教授。主要研究方向为智能发电、先进发电过程建模控制与优化。 E-mail: ffang@ncepu.edu.cn" ]
- 基金信息：
  
  国家重点研发计划(2025YFF0515900)
- DOI：10.12263/DZXB.20250999
  中图分类号： TP391.41;TK83
- 收稿：2026-01-28，
  
  录用：2026-02-24，
  
  纸质出版：2026-02-25
- 稿件说明：
移动端阅览
年嘉伟, 王梓懿, 唐灵犀, 等. Spatial-FineDef：融合多尺度感知与自适应增强的风电机组叶片小缺陷检测方法[J]. 电子学报, 2026, 54(02): 785-798.

NIAN Jiawei, WANG Ziyi, TANG Lingxi, et al. Spatial-FineDef: An Approach for Detecting Small Defects in Wind Turbine Blades that Integrate Multi-Scale Perception and Adaptive Enhancement[J]. Acta Electronica Sinica, 2026, 54(02): 785-798.
年嘉伟, 王梓懿, 唐灵犀, 等. Spatial-FineDef：融合多尺度感知与自适应增强的风电机组叶片小缺陷检测方法[J]. 电子学报, 2026, 54(02): 785-798. DOI：10.12263/DZXB.20250999

NIAN Jiawei, WANG Ziyi, TANG Lingxi, et al. Spatial-FineDef: An Approach for Detecting Small Defects in Wind Turbine Blades that Integrate Multi-Scale Perception and Adaptive Enhancement[J]. Acta Electronica Sinica, 2026, 54(02): 785-798. DOI：10.12263/DZXB.20250999

摘要

风电机组叶片是实现风能捕获与机械能转换的关键气动结构部件，其表面缺陷（如裂纹、侵蚀和脱胶）会影响叶片的气动性能进而降低机组的发电功率。实际风场巡检场景中，叶片表面缺陷通常存在缺陷微小、对比度低、纹理弱等特点，同时伴随复杂背景、光照变化和噪声侵蚀的干扰，使得目前的端到端检测方法在风电机组叶片小尺度缺陷检测任务中存在明显的性能受限问题。尽管实验室环境中的基准测试已取得超过99%的定位精度，但在实际场景中的复杂背景和小缺陷的影响下，定位与识别任务相耦合，降低了这类方法的检测精度。针对上述问题，本文提出一种融合多尺度感知与自适应增强的混合式两阶段检测方法Spatial-FineDef（Spatial-Fine Defect detection approach），通过解耦缺陷定位与精细判别过程，将缺陷候选区域提取与缺陷类别分类分阶段优化，提升了Spatial-FineDef在面对复杂背景和小缺陷问题时的检测精度。第一阶段，目标空间提取模块（Spatial-Net）通过结合定制化数据增强策略与定位增强方法提升了缺陷候选区域筛选的效率。第二阶段，精细化缺陷判别模块（FineDef-Net）在候选区域使用ConvNext主干结合轻量化多尺度注意力机制，在保持较低计算复杂度的前提下强化了精准区分不同缺陷的能力。相较于端到端式的故障检测方法，通过“先定位、后判别”的两阶段处理策略，Spatial-FineDef在抑制背景干扰的同时，实现了小缺陷区域的稳定筛选与精确类别判别。在现场采集的多缺陷风电机组叶片数据集中，Spatial-FineDef在麻面、涂层脱落、边缘开裂和表面裂纹四类缺陷检测任务中取得96.71%的整体准确率，并在多项指标上优于多种代表性基线模型。同时，消融实验验证了两阶段解耦策略与多尺度线性注意力模块在复杂背景下小缺陷检测的有效性。实验结果表明，本文提出的方法为风电机组叶片现场检测提供了一种可部署的技术方案，提升实时故障检测能力，并促进风电机组的可靠性及智能运维。

Abstract

Wind turbine blades are the core aerodynamic components responsible for wind energy capture and energy conversion in the wind turbine systems. Surface defects such as cracks

erosion

and delamination can deteriorate aerodynamic performance and consequently reduce power generation efficiency. In real-world wind farm inspection scenarios

blade surface detection is typically characterized with small scales and low contrast

and are often accompanied by complex backgrounds

illumination variation

and imaging noise. These challenges significantly limit the performance of existing end-to-end detection methods exhibit in small-scale defect detection tasks for wind turbine blades. Although the benchmark evaluations under laboratory conditions have achieved localization accuracies exceeding 99%

the strong coupling between localization and classification under strong background interference and micro-cracks makes these methods can hardly be further improved. To overcome these challenges

we propose Spatial-FineDef (Spatial-Fine Defect Detection Approach)

a hybrid two-stage detection method incorporating multi-scale perception and adaptive enhancement. Through explicitly decoupling defect localization from fine-grained classification

the proposed method optimizes candidate region extraction and defect recognition in a staged manner. In the first stage

Spatial-Net incorporates task-oriented data augmentation strategy and improved localization methods to enhance the accuracy of spatial filtering for potential defects. In the second stage

FineDef-Net utilizes ConvNext backbone with lightweight multi-scale linear attention mechanism which enhances the discrimination capability while maintaining low computational complexity. Compared with end-to-end detection methods

by adopting a two-stage strategy of localization followed by classification

Spatial-FineDef effectively suppresses background interference while enabling stable region selection and defect classification for small-scale defects. On a field-collected wind turbine blade fault dataset

Spatial-FineDef achieves an overall accuracy of 96.71% in the classification of four small-scale defect types including surface pitting

coating shedding

edge cracking

and surface cracking. Experimental results demonstrate outperforming multiple representative baseline models across several evaluation metrics. Ablation studies further validate the effectiveness of the decoupled two-stage strategy and the multi-scale linear attention mechanism in handling small defects under complex backgrounds. The proposed method provides a deployable and reliable solution for on-site blade inspection

facilitating real-time fault diagnosis and enhancing the reliability and intelligent operation and maintenance of wind turbine systems.

关键词

Keywords

references

朱海波 , 缪维跑 , 范士杰 , 等 . 大型风力机叶片仿生叶脉结构设计及性能分析 [J ] . 中国电机工程学报 , 2025 , 45 ( 4 ): 1521 - 1531 .

Zhu Haibo , Miao Weipao , Fan Shijie , et al . Design and performance analysis of bionic blade vein structure for the large wind turbine blade [J ] . Proceedings of the CSEE , 2025 , 45 ( 4 ): 1521 - 1531 . (in Chinese)

赵艳 , 杜艳强 , 王新武 , 等 . 风力机支撑结构疲劳损伤评估中的不确定性分析 [J ] . 中国电机工程学报 , 2023 , 43 ( 16 ): 6313 - 6319 .

Zhao Yan , Du Yanqiang , Wang Xinwu , et al . Uncertainty analysis in fatigue damage assessment of wind turbine support structure [J ] . Proceedings of the CSEE , 2023 , 43 ( 16 ): 6313 - 6319 . (in Chinese)

Wu Ping , Wang Yixuan , Zhang Xujie , et al . Wind turbine blade breakage monitoring with mogrifier LSTM autoencoder [J ] . IEEE Transactions on Instrumentation and Measurement , 2023 , 72 : 3534610 . DOI: 10.1109/tim.2023.3323967 http://dx.doi.org/10.1109/tim.2023.3323967

Encalada-Dávila Á , Moyón L , Tutivén C , et al . Early fault detection in the main bearing of wind turbines based on gated recurrent unit (GRU) neural networks and SCADA data [J ] . IEEE/ASME Transactions on Mechatronics , 2022 , 27 ( 6 ): 5583 - 5593 . DOI: 10.1109/tmech.2022.3185675 http://dx.doi.org/10.1109/tmech.2022.3185675

谢小荣 , 贺静波 , 毛航银 , 等 . “双高”电力系统稳定性的新问题及分类探讨 [J ] . 中国电机工程学报 , 2021 , 41 ( 2 ): 461 - 475 .

Xie Xiaorong , He Jingbo , Mao Hangyin , et al . New issues and classification of power system stability with high shares of renewables and power electronics [J ] . Proceedings of the CSEE , 2021 , 41 ( 2 ): 461 - 475 . (in Chinese)

Ran Xiukang , Zhang Shang , Wang Hengtao , et al . An improved algorithm for wind turbine blade defect detection [J ] . IEEE Access , 2022 , 10 : 122171 - 122181 . DOI: 10.1109/access.2022.3216613 http://dx.doi.org/10.1109/access.2022.3216613

Yuan Minghao , Zhou Yongbing , Ren Xiaoyu , et al . YOLO-HMC: An improved method for PCB surface defect detection [J ] . IEEE Transactions on Instrumentation and Measurement , 2024 , 73 : 2001611 . DOI: 10.1109/tim.2024.3351241 http://dx.doi.org/10.1109/tim.2024.3351241

Stokkeland M , Klausen K , Johansen T A . Autonomous visual navigation of Unmanned Aerial Vehicle for wind turbine inspection [C ] // 2015 International Conference on Unmanned Aircraft Systems . Piscataway : IEEE , 2015 : 998 - 1007 . DOI: 10.1109/icuas.2015.7152389 http://dx.doi.org/10.1109/icuas.2015.7152389

Wang Long , Zhang Zijun . Automatic detection of wind turbine blade surface cracks based on UAV-taken images [J ] . IEEE Transactions on Industrial Electronics , 2017 , 64 ( 9 ): 7293 - 7303 . DOI: 10.1109/tie.2017.2682037 http://dx.doi.org/10.1109/tie.2017.2682037

Deng Liwei , Guo Yan’gang , Chai Borong . Defect detection on a wind turbine blade based on digital image processing [J ] . Processes , 2021 , 9 ( 8 ): 1452 . DOI: 10.3390/pr9081452 http://dx.doi.org/10.3390/pr9081452

石腾 , 许波峰 , 陈鹏 , 等 . 基于机器视觉的风电机组叶片多类型损伤检测方法研究 [J ] . 太阳能学报 , 2024 , 45 ( 6 ): 487 - 494 .

Shi Teng , Xu Bofeng , Chen Peng , et al . Study on multi-type damage detection method for wind turbine blades based on machine vision technology [J ] . Acta Energiae Solaris Sinica , 2024 , 45 ( 6 ): 487 - 494 . (in Chinese)

Zhou Yan . A YOLO-NL object detector for real-time detection [J ] . Expert Systems with Applications , 2024 , 238 : 122256 . DOI: 10.1016/j.eswa.2023.122256 http://dx.doi.org/10.1016/j.eswa.2023.122256

Hang Xinyu , Zhu Xiaoxun , Gao Xiaoxia , et al . Study on crack monitoring method of wind turbine blade based on AI model: Integration of classification, detection, segmentation and fault level evaluation [J ] . Renewable Energy , 2024 , 224 : 120152 . DOI: 10.1016/j.renene.2024.120152 http://dx.doi.org/10.1016/j.renene.2024.120152

Ye Xingyu , Wang Long , Huang Chao , et al . Wind turbine blade defect detection with a semi-supervised deep learning framework [J ] . Engineering Applications of Artificial Intelligence , 2024 , 136 : 108908 . DOI: 10.1016/j.engappai.2024.108908 http://dx.doi.org/10.1016/j.engappai.2024.108908

Ma Limei , Jiang Xingyu , Tang Zizhen , et al . Wind turbine blade defect detection algorithm based on lightweight MES-YOLOv8n [J ] . IEEE Sensors Journal , 2024 , 24 ( 17 ): 28409 - 28418 . DOI: 10.1109/jsen.2024.3430351 http://dx.doi.org/10.1109/jsen.2024.3430351

Liu Lizhao , Li Pinrui , Wang Dahan , et al . A wind turbine damage detection algorithm designed based on YOLOv8 [J ] . Applied Soft Computing , 2024 , 154 : 111364 . DOI: 10.1016/j.asoc.2024.111364 http://dx.doi.org/10.1016/j.asoc.2024.111364

蒋兴群 , 刘波 , 宋力 , 等 . 基于改进YOLO-v3的风力机叶片表面损伤检测识别 [J ] . 太阳能学报 , 2023 , 44 ( 3 ): 212 - 217 .

Jiang Xingqun , Liu Bo , Song Li , et al . Surface damage detection and recognition of wind turbine blade based on improved YOLO-v3 [J ] . Acta Energiae Solaris Sinica , 2023 , 44 ( 3 ): 212 - 217 . (in Chinese)

Guo Jihong , Liu Chao , Cao Jinfeng , et al . Damage identification of wind turbine blades with deep convolutional neural networks [J ] . Renewable Energy , 2021 , 174 : 122 - 133 . DOI: 10.1016/j.renene.2021.04.040 http://dx.doi.org/10.1016/j.renene.2021.04.040

He Yan , Niu Xiaobo , Hao Chuanpeng , et al . An adaptive detection approach for multi-scale defects on wind turbine blade surface [J ] . Mechanical Systems and Signal Processing , 2024 , 219 : 111592 . DOI: 10.1016/j.ymssp.2024.111592 http://dx.doi.org/10.1016/j.ymssp.2024.111592

Yu Zhengxu , Shen Dong , Jin Zhongming , et al . Progressive transfer learning [J ] . IEEE Transactions on Image Processing , 2022 , 31 : 1340 - 1348 . DOI: 10.1109/tip.2022.3141258 http://dx.doi.org/10.1109/tip.2022.3141258

Han Honggui , Meng Yuan , Wu Xiaolong , et al . A transfer learning-based multimodal feature fusion model for bearing fault diagnosis [J ] . IEEE Transactions on Instrumentation and Measurement , 2025 , 74 : 3530413 . DOI: 10.1109/tim.2025.3558745 http://dx.doi.org/10.1109/tim.2025.3558745

Wang Long , Yang Jinxu , Huang Chao , et al . An improved U-Net model for segmenting wind turbines from UAV-taken images [J ] . IEEE Sensors Letters , 2022 , 6 ( 7 ): 6002404 . DOI: 10.1109/lsens.2022.3184521 http://dx.doi.org/10.1109/lsens.2022.3184521

Na Lan , Cai Baigen , Zhang Chongzhen , et al . A heterogeneous transfer learning method for fault prediction of railway track circuit [J ] . Engineering Applications of Artificial Intelligence , 2025 , 140 : 109740 . DOI: 10.1016/j.engappai.2024.109740 http://dx.doi.org/10.1016/j.engappai.2024.109740

普智勇 , 郭新毅 , 刘发炳 , 等 . 基于无损检测和迁移学习的风电机组主轴缺陷识别 [J ] . 机电工程 , 2025 , 42 ( 6 ): 1084 - 1092 .

Pu Zhiyong , Guo Xinyi , Liu Fabing , et al . Defect identification of wind turbine spindle based on nondestructive testing and transfer learning [J ] . Mechanical & Electrical Engineering Magazine , 2025 , 42 ( 6 ): 1084 - 1092 . (in Chinese)

Yang Xiyun , Zhang Yanfeng , Wei Lyu , et al . Image recognition of wind turbine blade damage based on a deep learning model with transfer learning and an ensemble learning classifier [J ] . Renewable Energy , 2021 , 163 : 386 - 397 . DOI: 10.1016/j.renene.2020.08.125 http://dx.doi.org/10.1016/j.renene.2020.08.125

Liu Yuhang , Zheng Yuqiao , Wei Tai , et al . Lightweight algorithm based on you only look once version 5 for multiple class defect detection on wind turbine blade surfaces [J ] . Engineering Applications of Artificial Intelligence , 2024 , 138 : 109422 . DOI: 10.1016/j.engappai.2024.109422 http://dx.doi.org/10.1016/j.engappai.2024.109422

Liu Dandan , Liu Mingjie . Wind turbine blades defect detection based on global and local attention with multi-feature fusion [J ] . Applied Soft Computing , 2025 , 185 : 113914 . DOI: 10.1016/j.asoc.2025.113914 http://dx.doi.org/10.1016/j.asoc.2025.113914

Liu Jinyan , Yan Longbin , Chen Jie . A coarse-to-fine object detection framework for high-resolution images with sparse objects [C ] // 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing . Piscataway : IEEE , 2021 : 1 - 6 . DOI: 10.1109/mlsp52302.2021.9596518 http://dx.doi.org/10.1109/mlsp52302.2021.9596518

Yuan Xiang , Cheng Gong , Yan Kebing , et al . Small object detection via coarse-to-fine proposal generation and imitation learning [C ] // 2023 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2023 : 6294 - 6304 . DOI: 10.1109/iccv51070.2023.00581 http://dx.doi.org/10.1109/iccv51070.2023.00581

Ali Shah R , Urmonov O , Kim H . Two-stage coarse-to-fine image anomaly segmentation and detection model [J ] . Image and Vision Computing , 2023 , 139 : 104817 . DOI: 10.1016/j.imavis.2023.104817 http://dx.doi.org/10.1016/j.imavis.2023.104817

Liang Yun , Hu Zhiguang , Huang Junjie , et al . ToCoAD: Two-stage contrastive learning for industrial anomaly detection [J ] . IEEE Transactions on Instrumentation and Measurement , 2025 , 74 : 5012009 . DOI: 10.1109/tim.2025.3545987 http://dx.doi.org/10.1109/tim.2025.3545987

Varghese R , M S . YOLOv8: A novel object detection algorithm with enhanced performance and robustness [C ] // 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems . Piscataway : IEEE , 2024 : 1 - 6 . DOI: 10.1109/adics58448.2024.10533619 http://dx.doi.org/10.1109/adics58448.2024.10533619

Liu Xinyu , Peng Houwen , Zheng Ningxin , et al . EfficientViT: Memory efficient vision transformer with cascaded group attention [C ] // 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2023 : 14420 - 14430 . DOI: 10.1109/cvpr52729.2023.01386 http://dx.doi.org/10.1109/cvpr52729.2023.01386

Wang Lei , He Yigang , Shao Kaixuan , et al . An unsupervised approach to wind turbine blade icing detection based on beta variational graph attention autoencoder [J ] . IEEE Transactions on Instrumentation and Measurement , 2024 , 73 : 2500912 . DOI: 10.1109/tim.2023.3286011 http://dx.doi.org/10.1109/tim.2023.3286011

雷涛 , 翟钰杰 , 许叶彤 , 等 . 基于边缘引导和动态可变形Transformer的遥感图像变化检测 [J ] . 电子学报 , 2024 , 52 ( 1 ): 107 - 117 .

Lei Tao , Zhai Yujie , Xu Yetong , et al . Edge guided and dynamically deformable transformer network for remote sensing images change detection [J ] . Acta Electronica Sinica , 2024 , 52 ( 1 ): 107 - 117 . (in Chinese)

Zhu Xiaoxun , Hang Xinyu , Gao Xiaoxia , et al . Research on crack detection method of wind turbine blade based on a deep learning method [J ] . Applied Energy , 2022 , 328 : 120241 . DOI: 10.1016/j.apenergy.2022.120241 http://dx.doi.org/10.1016/j.apenergy.2022.120241

Chen Zhendong , Feng Xuefeng , Liu Li , et al . Surface defect detection of industrial components based on vision [J ] . Scientific Reports , 2023 , 13 : 22136 . DOI: 10.1038/s41598-023-49359-9 http://dx.doi.org/10.1038/s41598-023-49359-9

Yun S , Ro Y . SHViT: Single-head vision transformer with memory efficient macro design [C ] // 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2024 : 5756 - 5767 . DOI: 10.1109/cvpr52733.2024.00550 http://dx.doi.org/10.1109/cvpr52733.2024.00550

Tian Yunjie , Ye Qixiang , Doermann D . YOLOv12: Attention-centric real-time object detectors [PP/OL ] . V1. arXiv ( 2025-02-18 )[ 2026-01-26 ] . https://doi.org/10.48550/arXiv.2502.12524 https://doi.org/10.48550/arXiv.2502.12524 .

He Kaiming , Zhang Xiangyu , Ren Shaoqing , et al . Deep residual learning for image recognition [C ] // 2016 IEEE Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2016 : 770 - 778 . DOI: 10.1109/cvpr.2016.90 http://dx.doi.org/10.1109/cvpr.2016.90

Shaker A , Maaz M , Rasheed H , et al . SwiftFormer: Efficient additive attention for transformer-based real-time mobile vision applications [C ] // 2023 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2023 : 17379 - 17390 . DOI: 10.1109/iccv51070.2023.01598 http://dx.doi.org/10.1109/iccv51070.2023.01598

Han Dongchen , Wang Ziyi , Xia Zhuofan , et al . Demystify Mamba in vision: A linear attention perspective [PP/OL ] . V1.arXiv ( 2024-05-25 )[ 2026-01-26 ] . https://doi.org/10.48550/arXiv.2405.16605 https://doi.org/10.48550/arXiv.2405.16605 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于区域内存模型的空指针引用缺陷检测

基于信息融合的区块链系统隐匿安全补丁识别及迁移技术

加密流量侧信道泄漏的不可避免性

基于超表面的低副瓣高口径效率反射阵列天线

基于注意力惩罚和自适应学习的场景图增强联合多模态方面情感分析