ATTRIBUTE EXTRACTION FOR AUTOMATIC CLASSIFICATION OF POLYMERIC INSULATORS BASED ON ULTRAVIOLET IMAGING

Summary

Polymeric insulators are more susceptible to damage caused by corona and surface discharges, when compared to ceramic insulators. These phenomena can cause reduction in mechanical strength, erosion of the polymer surface and creation of various chemical substances, which lead in the long term, to progressive deterioration and possible equipment failure. However, surface corona discharge and electrical field are also concentration indicators, and can provide good indicators of the insulating operating condition. In this work, a computational routine based on digital image processing was developed, which was obtained with the aid of a camera sensitive to ultraviolet radiation. The computer algorithm is capable of quantifying the number of discharges present in the video frame, areas of faculae and distances between faculae and the high voltage end fitting distance. Also, this routine is able to suggest the insulator division into specific regions, video frame divide into areas correlated to these regions according to the degree of importance relative to the spatial location of the detected discharges. Tests were performed with 230 kV polymeric insulators. Based on the distances, sizes, and persistence of discharges in respect to the insulator, which are the outputs of the computational routines, one has a better assessment of the level of degradation of the insulator, which helps efficient decision-taking for the classification of its operational status.

Additional informations

Authors

Zhang Shuqi, Wu Chao, Farzaneh Masoud, Kurrat Michael, Haake Frank