p-Index From 2014 - 2019
0.408
P-Index
This Author published in this journals
All Journal Jurnal Gaussian
Berta Elvionita Fitriani, Berta Elvionita
Unknown Affiliation

Published : 2 Documents
Articles

Found 2 Documents
Search

PERAMALAN BEBAN PEMAKAIAN LISTRIK JAWA TENGAH DAN DAERAH ISTIMEWA YOGYAKARTA DENGAN MENGGUNAKAN HYBRID AUTOREGRESIVE INTEGRATED MOVING AVERAGE – NEURAL NETWORK Fitriani, Berta Elvionita; Ispriyanti, Dwi; Prahutama, Alan
Jurnal Gaussian Vol 4, No 4 (2015): Wisuda periode Oktober 2015
Publisher : Jurusan Statistika UNDIP

Show Abstract | Original Source | Check in Google Scholar | Full PDF (705.302 KB)

Abstract

Excessive use of electronic devices in household and industry has made the demand of nation’s electrical power increase significantly these days. As a corporation that aim to provide national electrical power,  Perusahaan Listrik Negara (PLN) that distributes electrical power to Central Java and Yogyakarta has to be able to provide an economical and reliable system of electrical power provider. This study aimed to forecast data of electrical power usage in Central Java and Yogyakarta for the next 30 days. There were three forecasting methods used in this study; Neural Networks and Hybrid ARIMA-NN.  The data used in this study was electrical power usage data in January 2014 - November 2014 in Central Java and Yogyakarta. The accuracy of the study was measured based on MSE criteria where the best model chosen was the model that has lowest MSE value. According to the result of the analysis, using Neural Networks model to forecast electrical power usage for the next 30 days has better forecasting result than Hybrid ARIMA-NN model.Key Word : electrical power usage, forecasting of electrical power usage, ARIMA, NN, hybrid ARIMA-NN
PERAMALAN BEBAN PEMAKAIAN LISTRIK JAWA TENGAH DAN DAERAH ISTIMEWA YOGYAKARTA DENGAN MENGGUNAKAN HYBRID AUTOREGRESIVE INTEGRATED MOVING AVERAGE – NEURAL NETWORK Fitriani, Berta Elvionita; Ispriyanti, Dwi; Prahutama, Alan
Jurnal Gaussian Vol 4, No 4 (2015): Wisuda periode Oktober 2015
Publisher : Departemen Statistika FSM Undip

Show Abstract | Original Source | Check in Google Scholar | Full PDF (705.302 KB)

Abstract

Excessive use of electronic devices in household and industry has made the demand of nation’s electrical power increase significantly these days. As a corporation that aim to provide national electrical power,  Perusahaan Listrik Negara (PLN) that distributes electrical power to Central Java and Yogyakarta has to be able to provide an economical and reliable system of electrical power provider. This study aimed to forecast data of electrical power usage in Central Java and Yogyakarta for the next 30 days. There were three forecasting methods used in this study; Neural Networks and Hybrid ARIMA-NN.  The data used in this study was electrical power usage data in January 2014 - November 2014 in Central Java and Yogyakarta. The accuracy of the study was measured based on MSE criteria where the best model chosen was the model that has lowest MSE value. According to the result of the analysis, using Neural Networks model to forecast electrical power usage for the next 30 days has better forecasting result than Hybrid ARIMA-NN model.Key Word : electrical power usage, forecasting of electrical power usage, ARIMA, NN, hybrid ARIMA-NN