Diagnosis of power transformers through Frequency Response Analysis applying multiresolution analysis and pattern classification

Detection of abnormalities within transformers is a current field of study which aims to prevent future permanent damages. Among the automatic diagnostic methodologies there are some proposals, which make use of the Frequency Response Analysis (FRA) technique as a condition assessment criterion for its sensitivity to variations in the windings and the iron core. However, these methodologies are not completely robust due to the parameters that are used for its discrimination, which are generally statistics indicators. This paper shows a new diagnosis methodology which has been successfully applied in other fields based on the implementation of Wavelet multiresolution analysis and pattern recognition methods applied to FRA measurement records as a condition assessment tool.