基于深度学习算法的风电机组叶片开裂缺陷分析

为实现对风电机组叶片表面缺陷检测的智能化,该研究应用无人机技术、图像视觉技术和深度学习算法,建立风电机组叶片缺陷检测系统,提高了对叶片上开裂缺陷的检测精度;系统使用sobel算子计算图像横向和纵向的梯度,并对图像进行阈值分割和去噪处理;构建深度学习模型提取图像缺陷的特征信息,加入了SPP-Net网络进行卷积操作,增加了模型的输入数据尺度,得到特征图后在利用PRN网络筛选特征图;实验结果显示该研究系统能够去除大量无用的背景信息,开裂缺陷部位的特征信息保留完整,对验证集中的图像进行测试后,该研究系统识别出的开裂缺陷数最高可达到50个。

Crack Defect Analysis of Wind Turbine Blade Based on Deep Learning Algorithm

In order to realize intelligent detection of surface defects of wind turbine blades, this study applied UAV technology, image vision technology and deep learning algorithm to establish a defect detection system for wind turbine blades, which improved the detection accuracy of cracking defects on blades. The system uses an operator to calculate the horizontal and vertical gradients of the image, and performs threshold segmentation and denoising. The deep learning model was constructed to extract the feature information of image defects. Spp-net network was added for convolution operation, and the input data scale of the model was increased. After obtaining the feature image, PRN network was used to screen the feature image. The experimental results show that the system can remove a large amount of useless background information and retain the feature information of the crack defect position completely. After testing the images in the verification set, the number of crack defects identified by the system can reach 50 at most.