Surface Damage Detection Method for Blade of Wind Turbine Based on Image Segmentation

Bin Hu,Weibin Li,Chao Song,Keyi Yuan,Fan Zhao,Dong Wei

DOI: https://doi.org/10.1109/ccisp51026.2020.9273461

2020-01-01

Abstract:Wind energy, as a renewable energy source, has attracted great attention of many countries. It is converted into electricity by the important machine, wind turbine (WT). With the fast growing of WT capacity in the world, the costs of monitoring and maintaining increase year by year. One necessary part of these costs is for blades health situation monitoring, which directly affects effectiveness and safety of WT. Thus, the damage detection of the blades is of great significance. The blades suffer from sunshine, acid rain, sand attack, and others, which may cause kinds of damages, such as skin abscission, surface crack, and sand hole. Aiming at surface damage detection of the blades, the paper proposed a method based on variational image segmentation model. At beginning, an idea that objects in an image are left along with detection of the background part is present, and it is changed into an optimization problem by using the level set method and 1-D Heaviside function. Then the corresponding gradient descent minimization algorithm is derived. In the experimental part, a large number of blade images with surface damages are tested by applying the proposed method, and damages are all detected successfully. Moreover, the segmentation results of the noise images show that our method is robust to noise. This paper can provide effective technical support for surface damage detection of WT blades in-service.