Analysis of the Performance of the Moving Average Method in Forecasting the Increase in the Number of PLN Customers on Batam Island

Musfirah Dewi Lumentut (Electrical Engineering Department, Padang State Polytechnics, West Sumatera Province, Indonesia)
Khalil Rahman (Electrical Engineering Department, Padang State Polytechnics, West Sumatera Province, Indonesia)
Sarah Michelin Yunita (Chien Hsin University of Science and Technology, Taoyuan City, Taiwan)

Journal of Applied Mathematics and Modelling
Vol. 1 No. 1 (2025): Journal of Applied Mathematics and Modelling

Electricity demand is a crucial factor in regional development, requiring accurate forecasting to ensure efficient distribution. This study analyzes the performance of two forecasting methods, Simple Moving Average (SMA) and Weighted Moving Average (WMA), in predicting the increase in PLN customers on Batam Island for the period 2023-2024. The accuracy of these methods was evaluated using Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percent Error (MAPE), and performance indicators. The results indicate that both SMA and WMA provide accurate forecasts, with WMA showing slightly better accuracy due to its ability to assign different weights to past data. Based on the findings, the forecasted number of additional PLN customers in January 2025 using the WMA method is 386. This study demonstrates that the Moving Average method is a viable option for short-term electricity demand forecasting. However, its limitations in handling external factors suggest the need for more advanced forecasting models in future research.