Estimation of The Partial Discharge Position Using Neural Networks in The Power Transformers

신경망을 이용한 전력용 변압기의 부분방전 위치추정

This paper presents a new method for estimating partial discharge position using improved cross-correlation technique and neural networks in the power transformer. When ultrasonic signal is occurred by partial discharge, we detected these signals and calculated cross-correlation values with Hamming window. Also, we estimated partial discharge position using neural network with cross-correlation values. In the result of case study, we can estimate more accurately the partial discharge position than any other algorithms.