<![CDATA[Peak load forecasting plays a crucial role in power system operation planning by ensuring supply reliability and improving operational efficiency. This study aims to compare the forecasting performance of Linear Regression, Holt, and Holt–Winters methods in predicting the monthly peak load of PT PLN Batam and to identify the most accurate forecasting model. A quantitative time-series approach was employed using monthly peak load data collected from 2021 to 2025. Data from 2021–2024 were used for model training, while the 2025 dataset was reserved for validation. Model development was carried out in Python using the Pandas, NumPy, and Statsmodels libraries. Forecasting accuracy was evaluated using Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). The results demonstrate that the Holt–Winters method outperformed the Linear Regression and Holt models, achieving an MAE of 8.1, an MSE of 115.1, an RMSE of 10.7, and a MAPE of 1.13%. Based on the selected model, the projected monthly peak load for PT PLN Batam in 2026 ranges from 694.79 MW to 747.85 MW. These findings indicate that the Holt–Winters method effectively captures both trend and seasonal patterns, making it a reliable approach for supporting generation capacity planning and operational decision-making in electric power systems.]]>
Copyrights © 2026
