TENGKU REZA SUKA ALAQSA, -
(2025)
FORECASTING ELECTRICITY CONSUMPTION IN RIAU PROVINCE USING THE ARTIFICIAL NEURAL NETWORK (ANN) FEED FORWARD BACKPROPAGATION ALGORITHM FOR THE 2024-2027.
Indonesian Journal of Electronics, Electromedical Engineering, and Medical Informatics, 7 (1).
ISSN 2656-8624
Abstract
Electricity production in Riau Province fluctuates between surplus and deficit, as reported by the Central Statistics Agency. From a peak of 3,758.75 GWh in 2017, production fell to 525.19 GWh in 2019, mainly due to lack of investment in new power plants and dependence on external electricity supply. This study addresses these challenges by using the Artificial Neural Network (ANN) Feed Forward Backpropagation method to forecast electricity demand from 2024 to 2027. This study aims to analyze the accuracy of the prediction through the Mean Absolute Percentage Error (MAPE), evaluate electricity consumption projections, and calculate the annual growth rate. The gap in this study is the inclusion of previously ignored variables, namely the GRDP of Government Buildings and the number of Government Building customers. The methodology used is Artificial Neural Network Feed Forward Backpropagation inspired by the functioning of the human brain. In the training of data training, the MAPE was obtained at 4,315%. The electricity consumption prediction obtained is 8,679 GWh in 2024, 9,690 GWh in 2025, 10,959 GWh in 2026, and 12,681 GWh in 2027. The growth rate is also projected to increase, namely 5.67% from 2023 to 2024, 11.65% from 2024 to 2025, 13.10% from 2025 to 2026, and 15.71% from 2026 to 2027. The results of this study show the growing electricity demand in Riau Province over the forecast period, which emphasizing the need for better energy planning and targeted investments in power generation.
Actions (login required)
 |
View Item |
