Condition monitoring of electrical assets using digital IRT and AI technique

Abstract

In this paper, an advancement approach considering an infrared thermography methodology is taken into account for pronouncing and diagnosing the fault persisting in the electrical equipment. This technology is mainly focused on non-contact and non-destructive property. It is a fast and reliable technique to inspect system without any interruption. In the field of the electrified area, maintenance and reliability of transmission and distribution system are one of the most critical issue which mostly suffers from few problems like loose connection, corrosion, and unbalanced loads. The loose connection arise the sag and corrosion on wire produce the more corona loss. In this paper, non-invasive method is employed to monitor the temperature of zinc oxide (ZnO) surge arrester. Surge arrester is utilized to analyze the hot region and exercise the watershed transform for the image segmentation and hot color mapping. Detection of hot regions is resembled through dark red color. Monitoring of surge arrester leakage current (SALC) is the main consideration to solve out the problems through infra-red thermo-gram (IRT) and artificial intelligence (AI) techniques. Artificial neural network (ANN) techniques utilized monitoring the condition of arrester within input constraints; arrester temperature, ambient temperature and humidity. These constraints are implemented to find out leakage current. The proposed method detects the hotness, hot region of the ZnO arrester and a relationship between the thermal characteristic and leakage current of surge arrester for condition monitoring.