Study On Detecting Surface Defects Of Wind Turbine Blade Based On Machine Vision

As a clean and renewable environment-friendly energy,wind energy has been widely applied in various fields.For the purpose of maximizing the use of wind energy,wind farms are generally selected in open areas such as offshore,Gobi and mountainous areas.On the land wind power,such as the Gansu Jiuquan wind power base,the wind sand weather is frequent,the blade surface of the wind turbine is eroded by wind sand perennial,causing surface pits,scratches,cracks and other defects.If these defects are not repaired in time,it often leads to irreversibly irreversible safety.At present,the detection methods of the traditional blade surface defects mainly include tapping anvils,ground telescope observations,and artificial visual inspections.Many of the above methods have many drawbacks such as large blind zones,long time-consuming,labor-intensive,high-altitude operations,low efficiency,large power loss caused by shutdown,and many others..Therefore,based on the above problems,this paper proposes a fast and efficient method to detect blade surface defects based on the machine vision theory and image processing technology,and then provides technical support for the safe operation of the wind power structure.This paper focuses on the following aspects:(1)Presented the development status of wind turbines and machine vision theory at home and abroad,and focused on the influence of external environmental erosion on the surface of wind turbines under normal working conditions of wind turbines,resulting in the classification,parameters,and current defects of the surface.Current routine testing methods;(2)Based on the actual defect type and defect shape of the blade surface,determine the subsystem parameters of the machine vision to optimize the components of the image acquisition subsystem to construct the non-destructive testing system for the blade surface defects,and then set up the detection platform for the wind turbine blade.The surface of the test specimen is subjected to image information acquisition;(3)Based on the theory of digital image processing and the practical requirements of wind blade surface defect recognition,the best algorithms for image surface preprocessing,image segmentation and morphological processing to satisfy blade surface defect identification and parameter calculation are optimized in the image processing flow.Combined with the image quality evaluation function,the parameters of the algorithm are optimized,and finally an effective image processing algorithm for wind power blade surface defects are compiled;(4)Through the extraction of the defect features in the image after the digital image processing and the identification of the region description sub-parameters such as the length,width,perimeter,and area of the scratches,combined with the laboratory small-power wind turbine to complete the blade surface defects Feature extraction of defect images,calculation of defect area parameters and evaluation of blade status.