<![CDATA[This study aims to accurately forecast short-term electrical energy demand in a Jabodetabek industrial facility to support sustainable manufacturing and precise utility management. To address the highly dynamic and non-stationary nature of energy consumption, this research applied Holt’s Double Exponential Smoothing method. A 24-month historical dataset, spanning from January 2024 to December 2025, was analyzed to generate a six-month operational projection. Model performance was rigorously evaluated using Mean Absolute Deviation (MAD), Mean Square Error (MSE), and Mean Absolute Percentage Error (MAPE). The forecasting demonstrated outstanding predictive accuracy, yielding a MAPE of 1.26%, a MAD of 20,79, and an MSE of 690.40. The projection indicates a consistent linear escalation, with energy demand expected to reach 2.020,72 MWh by mid-2026.These findings confirm that Holt’s algorithm effectively captures underlying linear trends without seasonal behaviors. Integrating this precise mathematical model into standard procedures provides crucial early warning signals for capacity planning, mitigates utility shortages, and strongly supports the strategic mandate of green operations.]]>
Copyrights © 2026
