Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : Asian Journal of Applied Education

Forecasting Electricity Sales Using the Artificial Neural Network Backpropagation Method Yosi Febria Utami; Gumgum Darmawan; Resa Septiani Pontoh
Asian Journal of Applied Education (AJAE) Vol. 2 No. 4 (2023): October 2023
Publisher : PT FORMOSA CENDEKIA GLOBAL

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55927/ajae.v2i4.6589

Abstract

PT PLN operates in the field of providing electrical energy and one of its goals is to meet consumer needs for electrical energy now and in the future, as well as PLN UID West Java. The initial step is to estimate how much electricity will be sold in the future. For this reason, electricity sales forecasting is carried out which can be taken into consideration by PLN UID West Java in making decisions. This research uses monthly electricity sales data in West Java for the last ten years. This data is not linear and not stationary, so an alternative method is used, namely Artificial Neural Network Backpropagation. Forecasting produces the best network architecture 12-7-1 with a MAPE of 2.965%. This architectural model is used to forecast electricity sales in West Java until August 2024.