Wijaksono, Yohanes Andri (2024) Peramalan Peningkatan Kebutuhan Energi Listrik Terhadap Pembebanan Transformator Daya Di PT. PLN (Persero) ULP Indrapura Menggunakan Metode Jaringan Saraf Tiruan Backpropagationgusti. Masters thesis, Institute Technology Tenth Nopember Surabaya.
![]() |
Text
6032211210-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (3MB) | Request a copy |
Abstract
PT. PLN (Persero) merupakan perusahaan di bidang kelistrikan di Indonesia. Dari tahun ke tahun, pertumbuhan kebutuhan akan energi listrik menjadi sangat signifikan karena semakin meningkatnya peralatan-peralatan yang menggunakan sumber energi listrik. Kebutuhan energi listrik semakin meningkat dari tahun ke tahun, sehingga perlu dipersiapkan peralatan pendukungnya, dalam hal ini adalah trafo daya dan feeder. Adanya pengembangan dan pertumbuhan kebutuhan tenaga listrik, PT. PLN (Persero) memerlukan peramalan peningkatan kebutuhan energi. Data penelitian merupakan beban feeder dari gardu induk dan bersifat time series, sehingga penelitian ini menggunakan metode peramalan untuk memprediksi kebutuhan energi listrik di masa depan yaitu metode jaringan syaraf tiruan backpropagation. Menentukan arsitektur JST-BP yang sesuai untuk mendapatkan hasil peramalan terbaik dengan nilai error yang kecil. Pada penelitian ini menggunakan arsitektur JST-BP dengan 1 input, 30 neuron hidden layer, dan 1 output. Hasil penelitian menunjukkan bahwa metode peramalan kebutuhan listrik menggunakan JST-BP metode training conjugate gradient memiliki error yang kecil, yaitu pada Trafo Daya-1 MAPE Train= 1.5995 MAPE Test=1.842 MSE Train=0.489 MSE Test=0.423, Trafo Daya-2 MAPE Train= 2.8497 MAPE Test=1.6546 MSE Train=0.1214 MSE Test=0.072, Trafo Daya-3 MAPE Train=3.7452 MAPE Test=2.2556 MSE Train=0.8709 MSE Test=0.1721. Kemudian dilakukan prediksi hingga bulan Desember 2027, menunjukkan bahwa kebutuhan listrik sekitar 62% dari Kapasitas Transformator Daya, sehingga Transformator Daya 1, 2, dan 3 masih cukup mampu memenuhi kebutuhan listrik di PT. PLN(Persero) Unit Indrapura, dan untuk kemampuan hantar arus pada penyulang terdapat 2 penyulang yang berada di atas 80% dari KHA kabel, dan 8 penyulang di range 60% sampai dengan 80%, sehingga diperlukan adanya pembagian beban dengan uprating penampang kabel atau penambahan penyulang baru
===================================================================================================================================
PT. PLN (Persero) is a company in the electricity sector in Indonesia. From year to year, the growth in the need for electrical energy becomes very significant due to the increasing number of equipment that uses electrical energy sources. The need for electrical energy is increasing from year to year, so it is necessary to prepare supporting equipment, in this case the power transformer and feeder. With the development and growth of electricity demand, PT. PLN (Persero) requires forecasting of increased energy needs. The research data is the feeder load from the substation and is time series in nature, so this research uses a forecasting method to predict future electrical energy needs, namely the backpropagation artificial neural network method. Determine the appropriate ANN-BP architecture to get the best forecasting results with small error values. This research uses the ANN-BP architecture with 1 input, 30 hidden layer neurons, and 1 output. The results of the research show that the electricity demand forecasting method using JST-BP conjugate gradient training method has a small error, namely for Power Transformer-1 MAPE Train= 1.5995 MAPE Test=1.842 MSE Train=0.489 MSE Test=0.423, Power Transformer-2 MAPE Train = 2.8497 MAPE Test=1.6546 MSE Train=0.1214 MSE Test=0.072, Power Transformer-3 MAPE Train=3.7452 MAPE Test=2.2556 MSE Train=0.8709 MSE Test=0.1721. Then predictions were made until December 2027, showing that the electricity demand was around 62% of the Power Transformer Capacity, so that Power Transformers 1, 2, and 3 were still quite capable of meeting the electricity needs at PT. PLN(Persero) Indrapura Unit, and for the current carrying capacity of the feeder there are 2 feeders which are above 80% of the KHA of the cable, and 8 feeders in the range of 60% to 80%, so it is necessary to distribute the load by uprating the cable cross section or adding new feeder
Item Type: | Thesis (Masters) |
---|---|
Uncontrolled Keywords: | Tenaga Listrik, Neuron, Transformator, Penyulang, Peramalan, Jaringan Syaraf Tiruan, Backpropagation; Electric Power, Neuron, Transformer, Feeder, Forecasting, Artificial Neural Network, Backpropagation |
Subjects: | T Technology > T Technology (General) > T174 Technological forecasting |
Divisions: | Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT) |
Depositing User: | Yohanes Andri Wijaksono |
Date Deposited: | 13 Feb 2024 08:38 |
Last Modified: | 13 Feb 2024 08:38 |
URI: | http://repository.its.ac.id/id/eprint/107287 |
Actions (login required)
![]() |
View Item |