The Study On The Testing Of The Damaged Paint Coat Of The Blades Of Wind Turbine Generators With The Employment Of Infrared Thermal Imaging Technique

In the northern area of China,because of high frequency of the dry and windy weather,the paint coat of the blades of the wind turbine generator is vulnerable to the erosion of the sands in the wind.The erosion can causing the defects,such as tiny leak,rugged surface and cracks.If these defects can not be handled and restored,they can lead to major accidents,causing huge economic loss.The current testing method to the damage of the blades depends too much on the testing staffs' experience,which lacks the objectivity and science.In addition,this method has blind point and low efficiency.In consideration of these common defects of the paint-coat surface of the blades,the author of this paper proposes the testing method of infrared thermal imaging combined with digital picture processing technique of Matlab,wishing to lay theoretical root and technique support for the researches on blade-defect testing.This paper aims to studying the following aspects:(1)The author analyzes the blades' defect types and the location in the condition of operation,transportation and loading.(2)The 1.3-meter blades were employed,and the man-made defects are done to the surface paint coat by the erosion-and-abrasion experiment rig.Sequentially,infrared thermal imaging device of M7600E is used to collect the photographic images of the damaged blades.The author analyzes the impact of the cumulative time of thermal excitation and the distance of photographing on the sharpness of infrared thermal images.(3)Through the digital picture processing technique,the author enhances infrared thermal image of the damaged paint coat of the chosen blades by two-dimension discrete wavelet transform of wavedec2 functions.Afterwards,the images were processed in gray scale by weighed average algorithm.By changing the threshold value,the images were transformed to binary images which were then processed by erosion and inflation.By applying to Prewitt algorithm,the author conducted the edge detection and finally obtained the picture segmentation of the damaged blades.The results reveal that it is better to process and analyze the images of the damages caused by erosive wear after the cumulative time---irradiated by the sun---for 2 to 5 hours and taken in the distance of 0.8 to 1.0 meter.On the fair weather,the images of oil attachment is easily to be processed through the program.On the cloudy weather,it is hard to collect the infrared thermal images.Among all the man-made damages,the damages caused by erosive wear are more easily and clearly manifested in the infrared thermal image and their processing.