摘要
针对风力发电机叶片损伤人工检测效率低,受作业人员业务水平制约等因素影响大的问题,提出一种基于热红外图像的风力机叶片损伤识别以及损伤位置判定和损伤大小计算的方法,通过数字图像处理技术在Matlab 2019b平台开发实现。对附有人工损伤的风力机叶片热红外图像采用对比度拉伸、Gabor滤波、二值化阈值分割等方法提取损伤特征,叶片的磨损损伤利用质心确定其位置,裂纹损伤通过2个端点来表征损伤位置。最后根据磨损区域图像面积和裂纹长度,通过几何光学透镜成像法计算损伤大小。该方法实现了风力机叶片损伤的自动识别与测量,提高了风力机叶片损伤的检测效率。通过实验验证,该方法在风力机叶片损伤识别中具有较高的精确性。
A calculation method to identify the damage,damage position and size of the wind turbine blade based on thermal infrared image is put forward in this paper to tackle the problems of manual detection efficiency low and the large influence caused by operators.It has been developed and achieved on the Matlab 2019 b platform through digital image processing technology. It adopts the contrast stretch,Gabor filtering and binarization threshold segmentation to extract the damage features from the thermal infrared image of wind turbine blade with artificial damage,locates the position of abrasive damage through center of mass,and represents the damage position of cracks through two endpoints. The method makes the automatic identification and measurement of the damage for the wind turbine blade,which improves the detection efficiency of such damage. Experimental results show that this method has high accuracy in wind turbine blade damage identification.
作者
尹玉
张永
王健
吴国境
苏力德
王超誉
Yin Yu;Zhang Yong;Wang Jian;Wu Guojing;Su Lide;Wang Chaoyu(College of Mechanical and Electrical Engineering,Inner Mongolia Agricultural University,Hohhot 010018,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2022年第2期492-497,共6页
Acta Energiae Solaris Sinica
基金
草原英才“现代农牧业工程新技术研发及应用创新人才团队”(内组通字[2018] 19号)
内蒙古科技计划项目“风电机组叶片健康监测与维护关键技术研究”
内蒙古农业大学高层次人才引进科研启动项目(NDGCC2016-03)
中国博士后科学基金(2018M643777XB)。
关键词
风力机
叶片
红外热成像
图像处理
损伤识别
wind turbines
blades
infrared thermal imaging
image processing
damage identification
