Article Details

Abstract

To address the problems of low detection efficiency, poor recognition accuracy for small targets, and poor adaptability to engineering deployment in traditional wind turbine blade damage detection under complex environments, an intelligent detection scheme based on Deep Learning is proposed. First, a multi-scene dataset of wind turbine blades is collected, augmented, and annotated to construct a comprehensive dataset. Second, through a systematic and comprehensive comparison of five object detection algorithms—YOLOv7, YOLOv8, YOLOv9, YOLOv10, and YOLOv11—YOLOv11 is selected as the optimal detection model and fused with five mainstream attention mechanisms, including CBAM, ASPP, SKNet, ShuffleAttention, and UFO-ViT, to perform damage detection on the blades. Experimental results confirm that the YOLOv11+SKNet detection scheme significantly improves both the accuracy and efficiency of wind turbine blade damage detection in complex environments. Its mAP reaches 89.6%, a 0.6% improvement over YOLOv11, which effectively enhances the multi-scale feature extraction capability of the model and optimizes target localization accuracy. This method can provide reliable technical support and a theoretical basis for the intelligent operation and maintenance of wind power equipment.

Keywords

wind turbine blade; damage detection; YOLOv11; attention mechanism; SKNet

References

1. [1] Li Shuai Surface damage detection of wind turbine blades based on improved YOLO algorithm [D]. Dalian: Dalian Jiaotong University, 2024
2. [2] Yan Haowei Automatic detection method for surface defects of wind turbine blades based on unmanned aerial vehicles [J]. Computer Measurement and Control, 2024, 32 (11): 72-79
3. [3] Li Bing, Zhang Yimu, Wei Letao, etc A defect detection algorithm for wind turbine blades based on improved YOLOv8 [J]. Electronic Measurement Technology, 2024, 47 (13): 89-99
4. [4] Zeng Yongjie, Fan Bishuang, Yang Yawen, etc Application of Improved YOLOv8 Algorithm in Defect Detection of Wind Turbine Blades [J]. Journal of Electronic Measurement and Instrumentation, 2024, 38 (8): 26-35
5. [5] Liu Yi, Chu Yinhe, Li Bo Corrosion detection of fan blades based on continuous thermal excitation infrared detection [J]. Laser and Infrared, 2023, 53 (5): 716-722
6. [6] Zhao Juan, Chen Bin, Li Yongzhan, etc Acoustic feature extraction method for wind turbine blade crack faults under complex background noise [J]. Journal of Beijing University of Posts and Telecommunications, 2017, 40 (5): 117-122
7. [7] Marchevsky M，Prestemon S，Lobkis O，et al. Ultrasonic Waveguides for Quench Detection in HTS Magnets [J].IEEE Transactions on Applied Superconductivity，2022，32（6）：1-5.
8. [8] Tang Zhanjun, Zhang Chaojie, Wang Jian, etc Defect detection of fan blades based on improved YOLOv7 [J]. Journal of Beihang University, 2026, 52 (6): 1903-1914
9. [9] Wang Zhen Research and Implementation of Mobile Wind Turbine Blade Damage Identification Based on Deep Learning [D]. Beijing: North China Electric Power University (Beijing), 2023
10. [10] Wang C，Bochkovskiy A，Liao H，et al. YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition， 2023：7464-7475.
11. [11] Redmon J，Divvala S，Girshick R，et al. You Only Look Once：Unified，Real-Time Object Detection [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2016：779-788.
12. [12] Wang C，Yeh J，Liao H. YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information [C]//Proceedings of the 18th European Conference on Computer Vision，2024：1-18.
13. [13] Wang A，Chen H，Liu L，et al. YOLOv10: Real-Time End-to-End Object Detection [PP/OL].（2024-10-30）.https://arxiv.org/abs/2405.14458.
14. [14] Wang Haiqun, Guan Mei, Yu Haifeng, etc Improved YOLOv11n Fan Blade Defect Detection Algorithm [J/OL]. Electronic Measurement Technology, 2026:1-12 https://link.cnki.net/urlid/11.2175.tn.20260131.1515.006.
15. [15] Tang Zhanjun, Zhang Chaojie, Wang Jian, etc Defect detection of fan blades based on improved YOLOv7 [J]. Journal of Beihang University, 2026, 52 (6): 1903-1914
16. [16] Woo S，Park J，Lee J，et al. CBAM：Convolutional Block Attention Module [C]//Proceedings of the 15th European Conference on Computer Vision，2018：3-19.
17. [17] Chen L，Papandreou G，Kokkinos I，et al. DeepLab: Semantic Image Segmentation With Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs [J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2018，40（4）：834-848.
18. [18] Li X，Wang W，Hu X，et al. Selective Kernel Networks [C]//Proceedings of the 32nd IEEE/CVF Conference on Computer Vision and Pattern，2019：510-519.
19. [19] Zhang Q，Yang Y. SA-Net：Shuffle Attention for Deep Convolutional Neural Networks [C]//2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)，2021：2235-2239.
20. [20] He S，Guo P，Tang Z，et al. UFO-Net：A Linear Attention-Based Network for Point Cloud Classification [J].Sensors，2023，23（12）：5517.https://www.mdpi.com/1424-8220/23/12/5512.

Funding

First Class Discipline Research Special Project (YLXKZX-NGD-020); Basic Research Funds for Autonomous Region Directly Affiliated Universities (JY20250076)

About the authors

Lv Caixia (1990-), female, Han ethnicity, from Hohhot, Inner Mongolia, lecturer, master's degree, research direction: intelligent equipment manufacturing and algorithm design.