Hydrophobicity classification in high voltage polymeric insulators / Classificação da hidrofobicidade em isoladores elétricos poliméricos de alta tensão

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

The aim of this study is to propose a new method to classify the hydrophobicity (HC) in high voltage polymer insulators. Currently the HC is based on the STRI guide (Swedish Transmission Research Institute) published in 1992, where the hydrophobicity is classified by contact angle and/or amount of wet surface. In the proposed method, the HC was analyzed by digital image processing using the textures of the images obtained from electric insulators polymeric materials. These analysis were evaluated using mathematical morphology, fractal dimension, entropy, energy, variance and homogeneity. Simulated texture images (synthetic Images) about the surfaces of the insulators were created using a silicon plate and isopropyl alcohol and distilled water solutions in proportions from 0 to 100% in volume of alcohol (AIA). From these solutions, images of drops on the insulator surfaces were obtained to measure the contact angle; and surfaces images using the AIA solutions were obtained and then employed in the analysis of the texture patterns, which were used as basis for the analyzed methods. To analyze the images regardless the illumination conditions, which is a serious problem in natural weather, histogram equalization (EQU) and/or a White Top-Hat filter (WTH) was applied. This filter works as a high-pass filter, reducing the illumination gradient from the images obtained in natural conditions. Another way to avoid the gradient illumination was the segmentation technique by edge detection. The fractal dimension calculations in grayscale images were performed by the box-count method, with box size ranging from 3 to 11. Mathematical morphology was used to quantify the shapes in the images. The derivate of the volume from the opening images was analyzed to identify the elements in the image. The classification of the image texture by the entropy revealed not only the lowest dispersion of the results, but also the shortest time processing, presenting as an appropriated method to classify the hydrophobicity. Gama alterations in the images was done to observe the entropy values modifications as a function of the illumination effects in the images. The values shown the low influence of this parameter in the results. To evaluate the image quality regarding to size, zoom and resolution, different images were produced to observe the modification of these parameters and analyze its influence in the entropy values. From this study was possible to propose a mathematical function which relates the surface hidrophobicity and the texture patterns obtained by the AIA images. Hence, the hidrophobicity classification could be embedded in mobile devices, as example, cell phones, and performed in loco. A smartphone software was then developed to analyze the surface image of an insulator, producing the hidrophobicity classification value in about one minute.

ASSUNTO(S)

digital image processing electrical insulators isoladores elétricos processamento digital de imagens hidrophobicity hidrofobicidade

Documentos Relacionados