<![CDATA[This study examines electricity load, emphasizing the need for accurate prediction and optimal distribution. Utilizing artificial neural networks and the backpropagation algorithm, the research leverages data from BPS Kabupaten Cilacap and PT. PLN (Persero) UP3 Kabupaten Cilacap. Various configurations for hidden layer neurons, epochs, and learning rates are explored to determine the optimal network architecture for forecasting. The selected model, with specific criteria, demonstrates high accuracy during training (MSE: 0.00099999, MAPE: 5.44%, Regression: 0.98226) and testing (MSE: 0.0009493, MAPE: 3.99%, Regression: 0.90709) phases. The conclusion affirms the effectiveness of the Backpropagation ANN method in predicting electricity load in Kabupaten Cilacap for the period 2023-2030, meeting PLN's tolerance of ≤ 10% based on the MAPE criteria.]]>
Copyrights © 2024
