In electric power system the estimation of different waveforms are important in many cases. Transformer impulse fault current estimation helps to identify the type and detect the exact location of fault in winding. The estimation of short circuit currents is necessary while sizing and selecting the equipments and protective devices. In this proposed work, an attempt has been made for estimation of these non-linear currents by wavelet network (WN) approach. Since function estimation is the major field of application for WN, it is well suited for this purpose. A stochastic gradient type algorithm is adopted for training and the results are presented to show the potential of the proposed scheme in function estimation. It is found that the trained WN is capable of estimating non-linear functions accurately and efficiently. The WN results are compared with the results obtained from EMTP simulation and analytical method, which shows satisfactory agreement between the two. The performance of WN on estimation is also compared with feed forward Artificial Neural Network.
[1]
M. Moraud.
Wavelet Networks
,
2018,
Foundations of Wavelet Networks and Applications.
[2]
Y. C. Huang,et al.
Evolving Wavelet Networks for Power Transformer Condition Monitoring
,
2002,
IEEE Power Engineering Review.
[3]
C. B. Cooper,et al.
IEEE Recommended Practice for Electric Power Distribution for Industrial Plants
,
1987
.
[4]
Pasquale Daponte,et al.
Wavelet network-based detection and classification of transients
,
2001,
IEEE Trans. Instrum. Meas..
[5]
D. S. Crisford.
Introduction to IEC 909 (BS 7639). Short-circuit current calculation in three-phase AC systems
,
1996
.
[6]
Y.-C. Huang,et al.
Fault identification of power transformers using genetic-based wavelet networks
,
2003
.