Budiawan, Shafirah Khairina (2020) Simulasi Dan Pemodelan Arcing Seri Berbasis Neural Network Dengan Mempertimbangkan Beban Harmonik Dan Lokasi Arcing. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07111640000058-Undergraduate_Thesis.pdf Download (2MB) | Preview |
Abstract
Salah satu gangguan kelistrikan yang terjadi pada sistem tegangan rendah adalah arcing seri. Arcing seri terjadi ketika terdapat dua titik yang memiliki nilai potensial yang berbeda pada sebuah sambungan konduktor sefasa, atau umumnya pada kabel dengan isolasi terbuka atau terkelupas. Terjadinya arcing seri jarang disadari karena tidak terlihat dan ketika terjadi secara terus menerus maka temperatur pada daerah di sekitar titik arcing akan meningkat dan berpotensi menyebabkan kebakaran terjadi. Pada saat arcing seri terjadi arus gangguan memiliki nilai yang hampir sama dengan besar arus nominal sehingga peralatan pengaman seperti circuit breaker dan fuse tidak dapat bekerja. Agar peralatan proteksi mampu bekerja dengan baik dalam mendeteksi arcing seri dibutuhkan pemodelan gelombang arcing seri tegangan rendah. Pada penelitian ini, dilakukan pemodelan arcing seri tegangan rendah pada beban non-linear, yang memiliki nilai THDi, dan pada kondisi impedansi jaringan yang berbeda. Input pemodelan berupa nilai tegangan, arus, dan daya sesaat sebelum arcing terjadi dengan target berupa resistansi arcing. Pemodelan dilakukan dengan menggunakan jaringan saraf tiruan (artificial neural network) dengan algoritma feed-forward back-propagation. Berdasarkan penelitian yang dilakukan, didapatkan bahwa semakin besar nilai THDi maka semakin besar pula nilai Iarc, selain itu pemodelan yang dilakukan mampu merepresentasikan resistansi saat arcing terjadi dengan nilai MSE <0.04.
========================================================================================================================
Series Arc is one of the many type of electrical fault that occured in low voltage power system. Series arc occurred when two points of the same conductor connection have different potential values, it usually happens on a cable with broken insulation. The happening of series arc is rarely noticed by most people because it may not be visible, and when it happened continuously the temperature around the arc location will increase and may potentially cause fire. The series arc fault current has almost the same value as the nominal current, this caused protection devices such as circuit-breaker and fuse to not detect it. In order for protection devices to operate properly in detecting series arc fault, low voltage series arc modelling is needed. This experiment conducts a modelling of low voltage arc on non-linear loads, which contain THDi values, and on several different line impedance. The modelling input is the arc voltage, arc current, and arc power before series arc occurred and the modelling taget is the arc resistance. The modelling method is by using artificial neural network with feed-forward back propagation. Experiment shows that the higher the THDi values in the system, the higher the series arc fault current will be. The modelling results show that the modelling is able to representate the series arc fault resistance with MSE value less than 0.04.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | Pemodelan Arcing Seri, Tegangan Rendah, THDi, Impedansi Jaringan, dan Artificial Neural Network Series Arc Fault Modelling, Low Voltage, THDi, Line Impedance, Artificial Neural Network |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3226 Transients (Electricity). Electric power systems. Harmonics (Electric waves). T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7870.23 Reliability. Failures |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis |
| Depositing User: | Shafirah Khairina Budiawan |
| Date Deposited: | 07 Aug 2020 02:56 |
| Last Modified: | 18 May 2023 14:51 |
| URI: | http://repository.its.ac.id/id/eprint/77160 |
Actions (login required)
 |
View Item |