Research On Composite Insulator’s Hydrophobicity On-line Detection System Based On Image Processing

To prevent large-scale pollution flashover, the composite insulator, because of its hydrophobicity and hydrophobic migration characteristics, has been widely used in China. During operation, the composite insulator’s sheds is gradually aging by the integrated influence of ultraviolet light, pollution, electric field and so on. And its hydrophobicity is decreased or even lost as it’s getting old, increasing the probability of pollution flashover. To prevent pollution flashover that due to aging composite insulators, the aging degree of composite insulators should be detected regularly. The hydrophobicity of the composite insulator is directly related to its aging degree, so the aging grade of the composite insulator can be evaluated by detection of its hydrophobicity. At present, the domestic and foreign on-line composite insulator hydrophobicity detection technology is not mature. A composite insulator hydrophobicity on-line detection system is designed in this paper, which is composed of the system for wet composite insulator’s image on-line acquisition and the software for identification of composite insulator’s hydrophobicity grade. The system provides a new method for the on-line detection of hydrophobicity of composite insulators.In this paper, a set of on-line composite insulator’s image collecting system is designed based on the wireless communication technology, which consists of water spray device, image acquisition device, wireless remote control device and computer. The water spraying and image acquisition of the system is finished under the control of the staff under the transmission tower through wireless remote control device, and the image of composite insulator is saved in the computer. The system avoid the risk of high-voltage that might lead to staff when they spray to composite insulator and take photos of composite insulator on the tower. The safety and reliability of the system is verified through tests in laboratory, installation and tests on the transmission tower. The sharpness of images, collected by the system, is high enough for the following identification of composite insulator’s hydrophobicity.The enhancement and adaptive denoising method is proposed based on NSCT and Retinex theory in this paper. In this method, the contrast ratio of image can be effectively improved and contour information of drops is nearly completely retained while the noise of the image can be effectively removed. Then, based on the morphological method, the image, after preprocessing, is divided into two values (binary image), and the characteristic parameters are extracted from the image. In this paper, the SVM and multi fractal algorithm are combined to identify the hydrophobicity level and the accuracy of this method is 95%.Based on C++Visual 6.0 development platform, a set of hydrophobicity grade identification software was developed by MATLAB and VC++ hybrid programming. The software combines three methods, the subjective judgment, automatic identification and manual identification, to make it easy for user to select appropriate method and mutually verify the accuracy of the results, which draw by different method. The interface of the software is simple and beautiful, and the software is simple to operate and it can run stablely.