Identifikasi Jenis Partial Discharge pada Sambungan Kabel Tegangan Tinggi Menggunakan Artificial Neural Network Berbasis Ekstraksi Fitur Envelope dan Data Augmentasi

Prayoga, Gama (2026) Identifikasi Jenis Partial Discharge pada Sambungan Kabel Tegangan Tinggi Menggunakan Artificial Neural Network Berbasis Ekstraksi Fitur Envelope dan Data Augmentasi. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022241130-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy

Abstract

Sambungan kabel tegangan tinggi merupakan salah satu titik kritis pada sistem tenaga listrik yang rawan mengalami kerusakan akibat adanya degradasi isolasi. Fenomena aktivitas Partial Discharge (PD) sering muncul sebagai indikator awal dari penurunan kinerja isolasi. Adanya kontaminasi atau defect internal pada isolasi kabel dapat menjadi pemicu terjadinya PD. Keterbatasan jumlah data PD dan kompleksitas sinyal PD menjadi tantangan utama dalam proses identifikasi jenis kerusakan. Penelitian ini mengusulkan metode identifikasi berbasis ekstraksi sinyal PD bentuk envelope dengan teknik augmentasi data yang dikombinasikan dengan Artificial Neural Network (ANN) untuk klasifikasi dan identifikasi kerusakan pada sambungan kabel tegangan tinggi. Data PD diperoleh dengan membagi sinyal PD dalam domain fase menjadi beberapa segmen dan mengambil nilai maksimum pada setiap segmen sehingga diperoleh nilai numerik bentuk envelope yang informatif. Untuk mengatasi keterbatasan data, digunakan beberapa metode augmentasi, yaitu Deep Auto-Encoder Generative Adversarial Network (DAE-GAN), Gaussian Amplitude Scalling and Noise Injection (GAS-NI), dan Dynamic Time Warping Based Augmentation (DTW-BA) untuk menghasilkan data augmentasi dengan validasi yang dilakukan secara adaptif. Proses klasifikasi dilakukan menggunakan ANN dengan metode pelatihan dua tahap, yaitu pretraining dan fine-tuning. Evaluasi dilakukan menggunakan metode 5-fold cross-validation. Hasil penelitian menunjukkan bahwa metode yang diusulkan mampu meningkatkan kemampuan untuk klasifikasi dan identifikasi, terutama pada jenis defect akibat kontaminasi dengan akurasi tinggi, dengan akurasi mencapai 93,33%.
======================================================================================================================================
High-voltage cable joints are one of the critical points in the electric power system that are prone to damage due to insulation degradation. The phenomenon of Partial Discharge (PD) activity often appears as an early indicator of declining insulation performance. Contamination or internal defects in cable insulation can trigger PD. The limited amount of PD data and the complexity of PD signals are the major challenges in identifying the type of damage. This study proposes an identification method based on the extraction of envelope-shaped PD signals using data augmentation techniques combined with an Artificial Neural Network (ANN) for the classification and identification of damage in high-voltage cable joints. PD data was obtained by dividing the PD signal in the phase domain into several segments and taking the maximum value in each segment to obtain an informative numerical value representing the envelope signal. To address data limitations, several augmentation methods are used, namely Deep Auto-Encoder Generative Adversarial Network (DAE-GAN), Gaussian Amplitude Scaling and Noise Injection (GAS-NI), and Dynamic Time Warping-Based Augmentation (DTW-BA) to generate augmented data with adaptive validation. The classification process is performed using ANN with a two-stage training method: pre-training and fine-tuning. Evaluation was conducted using 5-fold cross-validation. The results of the study indicate that the proposed method is capable of improving classification and identification performance with high level of accuracy, particularly for contamination-related defects, with high accuracy reaching 93,33%.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Artificial Neural Network, Augmentasi, Partial Discharge, Sambungan Kabel, Sinyal Envelope, Augmentation, Cable Joint, Envelope Signal
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK201 Electric Power Transmission
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Gama Prayoga
Date Deposited:	31 Jul 2026 08:25
Last Modified:	31 Jul 2026 08:25
URI:	http://repository.its.ac.id/id/eprint/142453

Actions (login required)

View Item