Prediksi Tegangan Flashover pada Isolator Porselen 150 kV Terkontaminasi Alami Menggunakan Metoda Nonlinear Autoregressive Exogenous Neural Network (NARX-NN)

Mardini, Hasanah (2024) Prediksi Tegangan Flashover pada Isolator Porselen 150 kV Terkontaminasi Alami Menggunakan Metoda Nonlinear Autoregressive Exogenous Neural Network (NARX-NN). S2 thesis, Universitas Andalas.

	Text (Cover Abstrak) 1. Cover Abstrak.pdf - Published Version Download (1MB)
	Text (BAB I Pendahuluan) 2. BAB I Pendahuluan.pdf - Published Version Download (1MB)
	Text (BAB V Penutup) 3. BAB V Penutup.pdf - Published Version Download (1MB)
	Text (Daftar Pustaka) 4. Daftar Pustaka.pdf - Published Version Download (1MB)
	Text (Full Text - Tesis) Tesis - Prediksi Tegangan Flashover pada Isolator Porselen 150 kV Terkontaminasi Alami Menggunakan Metoda Nonlinear Autoregressive Exogenous Neural Network (NARX-NN).pdf - Published Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Contaminants on insulators can degrade their performance and lead to flashovers, which disrupt the reliability of electrical power systems. This research aims to develop a neural network model for predicting flashover voltage on high-voltage porcelain insulators that have been naturally contaminated. Experimental tests were conducted to collect data on parameters such as temperature, humidity, contaminant levels (ESDD and NSDD), and flashover voltage. A Nonlinear Autoregressive Exogenous (NARX) neural network model was used to predict flashover voltage based on the input parameters obtained from the tests. The results demonstrate that the NARX model has a high correlation coefficient of 1, indicating 100% accuracy and excellent prediction capability. This model has the potential to be used for determining flashover voltage on naturally contaminated insulators in a non-destructive manner, supporting preventive maintenance scheduling and ultimately improving the reliability of electrical power systems.

Actions (login required)

View Item