Ardila, Yunita (2016) Metode Hibrida ARIMA Dan Multilayer Perceptron Untuk Peramalan Jangka Pendek Konsumsi Listrik Di Jawa Timur - Hybrid ARIMA And Multilayer Perceptron For Short Term Forecasting Of Electricity Consumption In East Java. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
9114205301-Master thesis.pdf - Published Version Download (5MB) | Preview |
Abstract
Peningkatan akan kebutuhan listrik Indonesia diperkirakan akan terus
mengalamai pertumbuhan rata-rata 6,5% per tahun hingga tahun 2020. Oleh
karena itu PT. PLN (Persero) harus menyediakan energi listrik yang tepat sasaran
sesuai dengan kebutuhan konsumen untuk setiap subsistemnya. Energi listrik
sendiri adalah hasil dari perubahan energi mekanik yang tidak bisa disimpan,
sehingga jika terjadi kehilangan energi listrik yang tidak tersalurkan secara tepat
maka PT. PLN (Persero) akan mengalami kerugian. Maka perlu adanya
perencanaan sistem distribusi yang tepat untuk menyalurkan energi listrik, agar
energi yang tersalurkan tidak terbuang begitu saja. Tujuan dari penelitian yaitu
meramalkan konsumsi listrik jangka pendek untuk setiap subsistem di wilayah
Jawa Timur dengan metode Hibrida ARIMA dan Multilayer Perceptron (MLP).
Pemodelan dilakukan di subsistem Kediri pada jam 13:30, 18:30, dan
05:30. Subsistem Paiton dimodelkan pada jam 13:30, 18:30, dan 22:30. Subsistem
Ngimbang dimodelkan pada jam 13:30, 19:30, dan 22:30. Untuk subsistem Krian
pemodelan dilakukan pada jam 11:00, 15:00, dan 22:30. Subsistem Krian-Gresik
dimodelkan pada jam 14:00. 15:00, dan 22:30. Kriteria pemilihan model terbaik berdasarkan pada nilai MAPE data out sample. Dengan adanya peramalan ini
diharapkan juga akan mengoptimalkan kinerja pembangkit-pembangkit yang
mensuplai per subsistem distribusi listrik dalam proses load balancing yang sesuai
dengan konsumsi pelanggan agar tidak ada energi listrik yang terbuang sia-sia
atau bahkan terjadinya kelebihan kapasitas pemakaian.
========================================================================================================================
An increase in demand for electricity in Indonesia is expected to continue
to grow by average of 6,5% per year until 2020. Therefore, PT. PLN (Persero) had
to provide electrical energy that are effective in accordance with the needs of
consumers for each its subsystems. The electrical energy itself is the result of a
changed in the mechanical energy that can not be stored, so that if loss of
electrical energy is not channeled properly then PT. PLN (Persero) will suffer
losses. It is necessary to plan a proper distribution system to distribute electrical
energy, so that channeled energy won’t be wasted. The aim of this research is to
predict the short-term electricity consumption for each subsystem in East Java by
using Hybrid ARIMA and Multilayer Perceptron (MLP).
Kediri subsystem is modeled at 13:30, 18:30, and 05:30. Paiton subsystem
is modeled at 13:30, 18:30, and 22:30. Ngimbang subsystem is modeled at 13:30,
19:30, and 22:30. Krian is modeled at 11:00, 15:00, dan 22:30. Krian-Gresik
subsystem is modeled at 14:00. 15:00, and 22:30. Criteria for selection of the best
model is based on the MAPE value of out sample data. Given this forecast is also
expected to optimize the load balancing performance of power plants that supply
electricity distribution for each subsystem in accordance with customer
consumption so that no electrical energy is wasted or even the over capacity
usage. Result of the analysis showed that MLP method provides better accuracy
rate for electricity consumption forecasting in East Java based on peak load for
each subsystem compared with ARIMA and Hybrid ARIMA & MLP.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTMT 658.403 55 Ard m |
| Uncontrolled Keywords: | ARIMA, MAPE, Listrik, Ramalan Jangka Pendek, MLP, PT. PLN (Persero), ARIMA, MAPE, Electricity, Short Term Forecasting, MLP, PT. PLN (Persero) |
| Subjects: | H Social Sciences > H Social Sciences (General) > H61.4 Forecasting in the social sciences |
| Divisions: | 61101-Magister Management Technology |
| Depositing User: | ansi aflacha |
| Date Deposited: | 25 Nov 2019 03:53 |
| Last Modified: | 25 Nov 2019 03:53 |
| URI: | http://repository.its.ac.id/id/eprint/72011 |
Actions (login required)
 |
View Item |