Classification of Impulse Fault Patterns in Transformers Using Wavelet Network

Accurate identification of winding faults that may develop during impulse testing of transformer is of great importance for ensuring reliability and quality of power supply. In the case of fault the resulting winding current gets changed to a certain extent. The pattern of fault current depends on the type of fault and its location along the length of the winding. The proposed method is based on wavelet network, which can be proposed as an alternative to feed forward neural network for approximating complex non-linear function such as the impulse fault current waveform. The trained wavelet network is used for estimation of impulse fault current of various type and location. By analyzing the extracted weight parameters of the optimally trained wavelet network the second stage of network can classify the fault type and its location along the winding. Results of electromagnetic transient program based digital model of transformer show the ability of this approach for classification of various types of impulse faults.