The Remaining Life of Distribution Transformer Prediction by Using Neuro-Wavelet Method

The distribution transformer is one of the important equipment in delivering electricity to consumers. Apart from the normal use, fault conditions in the transformer can cause the life of the transformer to decrease being not optimal performance up to operating life limit. Therefore, it is very important to calculate the remaining life of the transformer. The steps taken are calculating the remaining life of the transformer using IEC 60076-7 and predicting the remaining life of the transformer using wavelet transform and back propagation neural network. The parameters required for this study are transformer current signal, loading, and transformer age. Measurement of current and temperature of distribution transformers in North Surabaya was conducted with a rating of 20 KV/ 380-220 V. Transformer current measurement has been processed using wavelet transforms to obtain detailed coefficients used to calculate energy values and power spectral density (PSD). Energy values, PSD, and transformer loading are training and testing data on the back propagation neural network. The expected output method is the prediction of the remaining life of the transformer.