Recognition of Power Transformer faults using Wavelet based Neural Networks

Recognition of Power-Transformer Protection is a veryimportant task for the power system operation.In this work, a hybrid of wavelet transform and neural network (WNN) approach is introduced for PTP (Power-Transformer Protection) events classification.The PT(Power Transformer) waveform is first decomposed by four levels Daubechies-4(db4) waveletanalysis and the decomposed waveforms then be processed by the NN for PTP event classification.(4) By utilizing the WNN, the PTP event recognition system can be implemented with minimum neurons and produces maximum attainment.Furthermore, this technique can accommodate maximum training patterns automaticallyreconsider the ANN system.The proposed approach is implemented in a simulation program to verify the validation and classification accuracy.