Deteksi Lokasi Gangguan Menggunakan Artificial Neural Network pada Jaringan Distribusi Dengan Relay Jarak

Redho, Francesco (2026) Deteksi Lokasi Gangguan Menggunakan Artificial Neural Network pada Jaringan Distribusi Dengan Relay Jarak. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022241072_Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (17MB) | Request a copy

Abstract

Sulitnya menemukan gangguan pada sistem distribusi 20kV dikarenakan jaringan distribusi mempunyai banyak percabangan menyebabkan penurunan keandalan sistem. Identifikasi lokasi gangguan secara cepat dan akurat menjadi aspek krusial. Relay jarak merupakan proteksi yang biasa digunakan untuk mendeteksi jarak gangguan pada sistem transmisi yang panjang, sehingga perlu penelitian lebih lanjut untuk diterapkan pada sistem distribusi dengan percabangan. Penelitian ini mengusulkan pendekatan berbasis Artificial Neural Network (ANN) untuk memprediksi jarak lokasi gangguan dan mengidentifikasi relay terdekat yang merasakan gangguan. Data gangguan diperoleh dari hasil simulasi sistem tenaga yang mencakup beberapa segmen saluran distribusi dan titik proteksi, model sistem dibangun pada aplikasi simulasi DIgSILENT PowerFactory. Model ANN dirancang untuk menghasilkan keluaran berupa jarak gangguan dari relay terdekat. Hasil pengujian pada penelitian ini menunjukkan bahwa metode yang diusulkan mampu memprediksi jarak gangguan dengan rata-rata akurasi 99,94% pada line utama dan percabangan. Dengan hasil akurasi ini dapat dikatakan relay jarak mampu bekerja dengan baik pada jaringan distribusi yang mempunyai beberapa percabangan dengan menggunakan ANN untuk prediksi jarak gangguan. Pendekatan ini diharapkan dapat menjadi solusi alternatif dalam pengembangan sistem proteksi adaptif berbasis kecerdasan buatan.
======================================================================================================================================
The difficulty in locating faults in 20kV distribution systems, due to the extensive branching of the network leads to a decrease in system reliability. Therefore, rapid and accurate fault location identification has become a crucial aspect. While distance relays are commonly used to detect fault distances in long transmission systems, further research is required for their application in branched distribution systems. This study proposes an approach based on Artificial Neural Network (ANN) to predict fault location distances and identify the nearest relay sensing the fault. Fault data were obtained from power system simulations covering several distribution line Segmens and protection points, with the system model developed using DIgSILENT PowerFactory software. The ANN model is designed to output the fault distance from the nearest relay. Based on the simulation results, it shows that the proposed method is capable of predicting fault distances with an average accuracy of 99,94% on the main line and on the branches. Based on these accuracy results, it can be concluded that distance relays integrated with ANN can perform effectively in distribution networks with multiple branches. This approach is expected to serve as an alternative solution in the development of AI-based adaptive protection systems.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Artificial Neural Network, Kehandalan Jaringan Distribusi, Relay Jarak, Artificial Neural Network, Distance Relay, Distribution Network Reliability
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3030 Electric power distribution systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Francesco Redho
Date Deposited:	01 Aug 2026 03:20
Last Modified:	01 Aug 2026 03:20
URI:	http://repository.its.ac.id/id/eprint/141244

Actions (login required)

View Item