AN INTELLIGENT TECHNIQUE FOR THE HEALTH ASSESSMENT OF POWER TRANSFORMER USING THERMAL IMAGING

Irshad, Zainul A. Jaffery, Nadeem Ahmad, and Ashwani K. Dubey

Keywords

Condition monitoring, image analysis, infrared imaging, powertransformer

Abstract

Fault diagnosis based on condition monitoring provides the basis for a reliable and cost-effective power trasnsformers operation. Thermal imaging and fuzzy control have already been proposed as a tool for automatic condition monitoring of transformer. In the majority of past research using thermal images, qualitative analysis has been used to assess the health of the transformer. These methods were not concerned more about the automatic detection and classification of hot regions of interest. In the present work, different segmentation methods for ROI detection are compared, and finally, a modified region growing (RG) technique is proposed for condition monitoring. Seed point selection and stopping rule for modified RG method for threshold temperature are carried out as per the specifications by National Electrical Testing Association. Instead of selecting shape-based features in most cases of literature, temperature-based features are used for fault classification. A fuzzy classifier is used to classify the severity of the fault. Failure effect and criticality analysis is done to find the fault location, a possible cause of fault, and actions to be taken.

Important Links:



Go Back