Loading...
Improvement of internal fault detection algorithms to reduce training time of back-propagation neural networks for transformer differential protection schemes
Author(s)
Bunjongjit, S.
Date Issued
January 1, 2012
Type
Article
Abstract
This paper presents an algorithm based on a combination of Discrete Wavelet Transforms (DWT) and back-propagation neural networks for detection and classification of internal faults in a two-winding three-phase transformer. Fault conditions of the transformer are simulated using Electromagnetic Transients Program (EMTP) in order to obtain current signals. The training process for the neural network and fault diagnosis decision are implemented on MATLAB. In addition, the initial number of neurons for the first hidden layer to decrease duration time of train process is taken into account. Various cases based on Thailand electricity transmission and distribution systems are studied to verify the validity of the proposed algorithm. A comparison between the proposed technique and conventional training is presented. The result is shown that the proposed technique is very effective in reduce training time and gives a satisfactory accuracy. © 2012 Praise Worthy Prize S.r.l. - All rights reserved.
Citation
International Review of Electrical Engineering, 7(5), 5598-5609, 2012