<![CDATA[This research is motivated by the use of electrical energy that is not fully in accordance with the Energy Consumption Intensity (IKE) standard, where there are still rooms with excessive or suboptimal energy use. The study aims to evaluate and predict electrical energy consumption in Building A of the Faculty of Teacher Training and Education, Sultan Ageng Tirtayasa University in order to improve energy efficiency. The method used is quantitative research with a comparative approach, namely comparing manual calculations based on the IKE standard with the Adaptive Neuro Fuzzy Inference System (ANFIS) method. Data were obtained through observation and measurement of electrical power, room area, and duration of use which were then analyzed using Matlab with the stages of fuzzification, FIS formation, hybrid learning training, and evaluation using RMSE. The results showed that the ANFIS method produced a better level of accuracy than manual calculations. The lowest error value was obtained in the gbellmf membership function for training at 0.53 and gaussmf for testing at 0.45. These findings indicate that ANFIS is able to model nonlinear relationships effectively and has the potential to be applied as a support system for evaluating electrical energy efficiency]]>
Copyrights © 2026
