Model Prediksi Apparent Age Trafo Tenaga Pada Industri Kelistrikan

Kurniawan, Aloysius Kohan (2024) Model Prediksi Apparent Age Trafo Tenaga Pada Industri Kelistrikan. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6010221018-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy

Abstract

Penurunan kondisi peralatan merupakan fenomena alami yang umum terjadi. Tingkat penuaan peralatan yang bervariasi disebabkan oleh usia peralatan itu sendiri dan kondisi operasi peralatan. Apparent age merupakan kombinasi dari kondisi peralatan dan usia peralatan. Berbeda dengan penelitian yang telah dilakukan sebelumnya, pada penelitian ini perhitungan apparent age dilakukan dengan mengkombinasikan umur peralatan dengan data primer hasil pengujian trafo daya. Data primer untuk pengujian trafo menggunakan 13 parameter pengujian yang dikelompokkan dalam 3 faktor yaitu Oil Quality Factor (OQF), Paper Condition Factor (PCF) dan Dissolved Gas Analysis Factor (DGAF). Perhitungan Apparent Age dilakukan dengan 2 metode: Multiple Linear Regression (MLR) & Support Vector Regression (SVR). Hasil perhitungan dari kedua metode tersebut akan dibandingkan dan dipilih metode terbaik yang dapat merepresentasikan apparent age trafo tenaga. Dalam penelitian ini kriteria performansi terbaik diperoleh dengan model prediksi menggunakan metode SVR kernel RBF C 200 dan gamma 0,2 dengan performa: adjusted koefisien determinasi 0,9216, koefisien korelasi 0,9607, dan MAE 1,676 tahun. Model prediksi menghasilkan 19,1% trafo daya mengalami laju penuaan tinggi, 66,0% trafo daya mengalami laju penuaan sedang, dan 14,9% trafo daya mengalami laju penuaan rendah. Untuk mengoptimalkan sumber daya perusahaan (tenaga, waktu, biaya, risiko), work order inspeksi level 3 (shutdown maintenance) trafo tenaga diterbitkan untuk trafo tenaga yang masuk kedalam daftar trafo tenaga dengan laju penuaan tinggi.
=================================================================================================================================
Equipment Aging is a common natural phenomenon. Equipment aging rate vary by age of equipment itself and operating conditions of equipment. Apparent age is a combination of equipment condition and equipment age. Contrast with previous research, in this study apparent age calculated by combining equipment age with primary data from power transformer testing results. Primary data for transformer testing uses 13 test parameters which are collected in 3 factors: Oil Quality Factor (OQF), Paper Condition Factor (PCF) and Dissolved Gas Analysis Factor (DGAF). Apparent age calculated using 2 methods: Multiple Linear Regression (MLR) & Support Vector Regression (SVR). Calculation results of the two methods will be compared and the best method that can represent apparent age of power transformer will be selected. This study resulted that best performance carried out by SVR with RBF kernel C 200 and gamma 0,2 with the performance: adjusted determination coefficient 0,9216, coefficient correlation 0,9607, and MAE 1,676. Prediction model resulted that 19.1% power transformers experienced accelerated aging, 66.0% power transformers experienced normal aging and 14.9% power transformers experienced low aging rate. To optimize company resources (energy, time, costs, risks), power transformers maintenance work order just issued for power transformers that experienced accelerated aging.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Apparent Age, MLR, SVR, Trafo Tenaga, Power Transformer
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2551 Electric transformers.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26101-(S2) Master Thesis
Depositing User:	Aloysius Kohan Kurniawan
Date Deposited:	05 Feb 2024 08:12
Last Modified:	05 Feb 2024 08:12
URI:	http://repository.its.ac.id/id/eprint/106164

Actions (login required)

View Item