Rekonfigurasi Jaring Distribusi Berbasis Deep Learning

Nursyahbani, Iqbal (2024) Rekonfigurasi Jaring Distribusi Berbasis Deep Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022201085-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (11MB) | Request a copy

Abstract

Jaringan distribusi adalah jaringan yang sangat berhubungan dengan pelanggan. Namun, jaringan distribusi ini sering mengalami sejumlah masalah. Salah satunya adalah jaringan distribusi masih mengalami rugi daya yang signifikan, yang mengurangi keandalan sistem. Rekonfigurasi jaringan distribusi adalah cara mengatur ulang konfigurasi jaringan dengan membuka dan menutup switch jaringan distribusi. Ini dilakukan untuk mengurangi rugi daya jaringan distribusi dan meningkatkan keandalan sistem distribusi. Akibatnya, efisiensi daya yang disalurkan meningkat dan konsumen dapat dilayani dengan lebih baik. Dikarenakan data yang cukup banyak, pada tugas akhir ini digunakan sebuah sistem deep learning yaitu Extra Trees Classifier. Tujuan digunakan deep learning ini bukan hanya untuk melakukan optimasi pada jaringan distribusi, tetapi juga untuk mempercepat waktu perhitungan untuk menemukan solusi terbaik pada rekonfigurasi. Hasilnya, dari simulasi yang dilakukan dengan Extra Trees Classifier didapatkan prediksi switch open dan close yang sesuai dengan data aktual yang digunakan. Untuk waktu yang dibutuhkan untuk prediksi dengan input data training adalah 13.246932 detik dengan setiap input rata-rata memakan waktu 0.018424 detik.
===============================================================================================================
Distribution networks are networks that are closely related to customers. However, these distribution networks often experience a number of problems. One of them is that the distribution network still experiences significant power losses, which reduce system efficiency. Distribution network reconfiguration is a way to reset the network configuration by opening and closing the distribution network switch. This is done to reduce distribution network power losses and improve the efficiency of the distribution system. As a result, the efficiency of the power distributed increases and consumers can be served better. Due to the large amount of data, this final project uses a deep learning system, namely Extra Trees Classifier. The purpose of using this deep learning is not only to optimize the distribution network, but also to speed up the calculation time to find the best solution for reconfiguration. As a result, from the simulation carried out with Extra Trees Classifier, predictions of open and close switches were obtained that were in accordance with the actual data used. The time required for prediction with training data input was 13.246932 seconds with each input taking an average of 0.018424 seconds.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Reconfiguration, Switch, Deep learning, Extra Trees Classifier, Optimize, Rekonfigurasi, Switch, Deep Learning, Extra Trees Classifier, Optimasi
Subjects:	T Technology > T Technology (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK1322.6 Electric power-plants T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3030 Electric power distribution systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Iqbal Nursyahbani
Date Deposited:	01 Aug 2024 03:55
Last Modified:	01 Aug 2024 03:55
URI:	http://repository.its.ac.id/id/eprint/111688

Actions (login required)

View Item