ANALISIS RE-SETTING DISTANCE RELAY MENGGUNAKAN METODE ARTIFICIAL NEURAL NETWORK PADA SALURAN TRANSMISI SISTEM 150 KV GARDU INDUK BANGIL – GARDU INDUK PIER

Abstract

The results of conducting impedance measurements at the Bangil - PIER Substation using the Omicron CPC100 tool show the difference in the results with the factory specification impedance value which exceeds the tolerance limit of 5% of the test results. Re-setting the distance relay is needed so that the relay is able to properly secure the transmission line. In this study, the Artificial Neural Network model was used to re-setting the distance relay. Re-setting was carried out using Analysis Software Based on Modeling and Visualization showed that the relay worked very well in responding to disturbances without overlapping it ranges. The results of calculation using ANN showing a very high level of accuracy of 99.88% and an error rate of 0.00113. Other data analysis indicators used produce the following values: MSE = 0,00002, RMSE = 0,0047 , MAD = 0,00361 , and MAPE = 0,163% which indicates that the Neural Network model that has been made is feasible to use.