Laboratory prediction energy control system based on artificial intelligence network

Desmira Desmira,Norazhar Abu Bakar,Mohd Ruzaini Hashim,Romi Wiryadinata,Mustofa Abi Hamid
DOI: https://doi.org/10.11591/eei.v11i3.1821
2022-06-01
Bulletin of Electrical Engineering and Informatics
Abstract:The use of electrical energy increases globally every year. The laboratory prediction energy control system (LPECS) predicted energy demand. This research was conducted in the Electrical Engineering Vocational Education laboratory by comparing the artificial neural fuzzy system (ANFIS) with the fuzzy logic. The comparison of methods aimed to determine their reliability in the energy demand prediction systems. The results showed that the minimum value of the target data using the conventional method (actual data) was 44.42%. Meanwhile, the prediction data using the ANFIS method was 44.33%, and the prediction data using the fuzzy method was 59.31%. The maximum value of the conventional ways (actual data) of ANFIS and fuzzy was similar by 77.59%. The RMSE ANFIS value was 0.1355%, the mean absolute percentage error (MAPE) was 0.2791%, and the fuzzy logic was 0.1986%. Thus, the ANFIS is applicable to determine the minimum and maximum values. Meanwhile, fuzzy can only show the maximum value but cannot reach the minimum value properly.
What problem does this paper attempt to address?