New Approach Insulator Failure Differentiation based on Image Processing

Abstract

Maintenance planning is crucial to improve the quality of the power systems. Several non-destructive methodologies have been developed to enhance the reliability of power transmission and distribution. In this context, insulators play a key role in electrical systems since they allow the connection of the power grid to electrical machines, cables, transformers, etc. Among several flaws, the most common incipient failures in the insulators are characterized by poor contacts and partial discharges. These unconformities can lead the insulator to total failure and impair the power transmission. Once each type of failure requires different maintenance actions, this article presents a new approach to perform the differentiation between poor contact fault and partial discharge activity using thermal analysis. Besides the classification, this work evaluated the evolution of the faults. Several tests were carried out, and the two types of flaws were applied to an insulator. Thermal images were captured, and the new TSTD (Temperature Standard Deviation) index was used to achieve the classification.