Keywords
Artificial Neural Network
Backpropagation
PT. PLN
Abstract
The escalating demand for electrical energy necessitates accurate forecasting by PT. PLN (Persero), the primary electricity provider. This study focuses on analyzing electricity consumption trends at PT. PLN (Persero) UP3 South Makassar from 2023 to 2025 using the Artificial Neural Network Backpropagation method through Matlab R2020a. Experiments were conducted to ascertain the optimal model architecture by iterative testing of various configurations. The best outcomes were achieved utilizing six input variables, encompassing energy consumption data from the social, household, business, industrial, public, and special service sectors in South Makassar spanning 2017 to 2022. The hidden layer comprised 18 neurons, while the output layer featured a single neuron representing electricity consumption in MWh for 2023 to 2025. Forecasting results exhibited a high level of precision with an average Mean Absolute Percentage Error (% MAPE) of 10.47% and a Mean Squared Error (MSE) value of 1.0742. This remarkably accurate forecast encompassed an error margin of approximately 10%-20%. Examination of neural network training yielded the best training performance graph and a correlation approaching 1 between output and target. These prognostications furnish a robust foundation for PT. PLN (Persero) in formulating prudent policies to meet the anticipated surge in energy demand. This is predicated on historical data spanning 2017 to 2022.Keywords:References
Rosadi, M., & Syamsul Amar B. (2019). Faktor-Faktor yang Mempengaruhi Konsumsi Listrik di Indonesia. Jurnal Kajian Ekonomi dan Pembangunan, 1(2), 273-286.
Harahap, F. (2019). ”Skripsi Analisa Pengujian Prakiraan Kebutuhan Energi Listrik Wilayah Provinsi Sumatera Utara menggunakan Metode Backpropagation Neural Network.”
Sumijan et al. (2016) ‘Implementation of Neural Networks in Predicting the Understanding Level of Students Subject’, International Journal of Software Engineering and Its Applications, 10(10), pp. 189–204. Available at: http://dx.doi.org/10.14257/ijseia.2016.10.10.18.
Budiharjo et al. (2018a) “Predicting School Participation in Indonesia using Back-Propagation Algorithm Model,” International Journal of Control and Automation, 11(11), hal. 57–68.
Budiharjo et al. (2018b) “Predicting tuition fee payment problem using backpropagation neural network model,” International Journal of Advanced Science and Technology, 120, hal. 85–96. doi:
14257/ijast.2018.120.07.
https://web.pln.co.id/statics/uploads/2023/05/Statistik-PLN-2022-Final-2.pdf
https://wikielektronika.com/rumus-daya-listrik/2/
Pangestu, M., Andriawan, A., & Wardah, I. (2022). Artificial Neural Network untuk Memprediksi Konsumsi Energi Listrik di PT.PLN (Persero) UP3 Surabaya Selatan.
Nur Maulana, F. (2019). ”Skripsi Analisa Perbandingan Metode Single Exponential Smoothing and Holt-Winter Exponential Smoothing.”
Setyowati, D., & Sunardiyo S. (2020) “Jurnal Prakiraan Kebutuhan Energi Listrik dengan Jaringan Saraf Tiruan (Artificial Neural Network) Metode Backpropagation Tahun 2020-2025.”