Defect detection of wind turbine blade based on improved SSD

doi:10.19936/j.cnki.2096-8000.20220328.005

Abstract

Abstract: Aiming at the problems of low accuracy and time-consuming of traditional algorithm, this paper proposed a defect detection method of wind turbine(WT) blade based on improved SSD. Firstly, the inclined WT blades are corrected by Hough transform and trimmed to a uniform size to form a complete data set. Then, two kinds of residual networks (ResNet, ResNext) are used to replace the traditional VGG as the backbone of SSD algorithm for feature extraction. Finally, the prediction network is used to complete the prediction of defect category and boundary box location. The experimental results show that the improved SSD algorithm improves the value of mAP by 7% compared with the traditional SSD algorithm. At the same time, the detection efficiency has been greatly improved.

Key words: wind turbine blade, object detection, improved SSD, fault detection, composites

CLC Number:

TB332

ZHU Jia-wei, WEN Chuan-bo. Defect detection of wind turbine blade based on improved SSD[J]. COMPOSITES SCIENCE AND ENGINEERING, 2022, 0(3): 38-44.

References

[1] 毛希玮,徐莹莹. 基于无人机的风机叶片缺陷自动检测技术[J]. 复合材料科学与工程,2020(9): 85-89.
[2] ZABIHOLLAH A, ENTESARI F, ALIMOHMMADI H. Fault detection of wind turbine blade under sudden change of wind speed condition using fiber optics[C]//Preceedings of the 2015 IEEE Sensors Applications Symposium (SAS). Zadar: 2015: 1-4.
[3] 张保钦, 雷保珍, 赵林惠, 等. 风机叶片故障预测的振动方法研究[J]. 电子测量与仪器学报, 2014, 28(3): 285-291.
[4] OH K Y, PARK J Y, LEE J S, et al. A novel method and its field tests for monitoring and diagnosing blade health for wind turbines[J]. IEEE Transactions on Instrumentation & Measurement, 2015, 64(6): 1726-1733.
[5] WALLE G, ABUHAMAD M. TOMA E, et al. Defect indications in sono-thermography in relation to defect location and structure[C]//2004 Quantitative InfraRed Thermography. 2004: 1-6.
[6] GIRSHICK R, DONAHUE J, DARRELLT T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Washington, DC: IEEE Computer Society, 2014: 580-587.
[7] GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Washington, DC: IEEE Computer Society, 2015: 1440-1448.
[8] REN S, HE K, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis ＆ Machine Intelligence, 2017, 39(6): 1137-1149.
[9] HE K, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2017: 2961-2969.
[10] RESMON 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 (CVPR). Las Vegas, NV: 2016: 779-788.
[11] LIU W, AGUELOV D, ERHAN D, et al. SSD: Single shot multibox detector[C]//Proceedings of the 2016 European Conference on Computer Vision(CVPR). Berlin: Springer, 2016: 21-37.
[12] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014.
[13] 杨珺, 张闯, 孙秋野, 等. 风电场选址综述[J]. 太阳能学报, 2012, 33(s1): 136-144.
[14] 吴林阳, 杜伟健, 陈小兵, 等. 一种运算和数据协同优化的深度学习编译框架[J]. 高技术通讯, 2020, 30(2): 120-125.
[15] 朱晓慧, 钱丽萍, 傅伟. 图像数据增强技术研究综述[J]. 软件导刊, 2021, 20(5): 230-236.
[16] BALLARD D H. Generalizing the Hough transform to detect arbiyrary shapes[J]. Pattern Recogniton, 1981, 13(2): 111-122.
[17] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). USA: 2016: 770-778.
[18] XIE S, GIRSHICK R, HE K, et al. Aggregated residual transformations for deep neural networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2017: 5987-
5995.