Artificial neural network-based prediction technique for transformer oil breakdown voltage

This paper presents an artificial neural network (ANN)-based modeling technique for prediction of transformer oil breakdown voltage. This model comprises transformer oil service period, total acidity and water content while preserving the nonlinear relationship between their combinations for predicting transformer oil breakdown voltage. The model results are compared with those obtained by various modeling techniques such as ANN-based model for transformer oil breakdown voltage as a function of its set-vice period, a polynomial regression model for transformer oil breakdown voltage as a function of its service period and a multiple linear regression model for transformer oil breakdown voltage as a function of its total acidity, water content and service period. A quantitative analysis of various modeling techniques has been carried out using different evaluation indices; namely, mean absolute percentage error and actual percentage error at each service period. The results showed the effectiveness and capability of the proposed ANN-based modeling technique to predict transformer oil breakdown voltage and justified its accuracy. (C) 2004 Elsevier B,V. All rights reserved.