<![CDATA[The Early Warning System (EWS) Trafo is a continuous monitoring tool for transformers that provides warnings when anomalies are detected, aiming to prevent explosions. This device applies artificial intelligence and machine learning technologies to monitor and predict the real-time condition of transformers using sensor data collected by the tool. This research aims to predict the condition of transformers based on the EWS Trafo sensor results using the Extreme Learning Machine (ELM) method. The study investigates the effectiveness of the ELM method in predicting transformer conditions. Based on the research results obtained from several combinations of data training: testing with different numbers of hidden layers, the lowest Mean Absolute Percentage Error (MAPE) value was found in the combination of 40% training data and 60% testing data, out of a total of 470 data points, with 20 hidden layers, at 23.1125%. Thus, it can be concluded that the Extreme Learning Machine (ELM) method is effective in predicting the condition of transformers.]]>
Copyrights © 2024
