Widodo, Muhammad Budi Rahayu (2020) Analisis Deteksi Lokasi Gangguan Pada Sistem Distribusi 20 Kv Menggunakan Injeksi Sinyal Impulse Yang Dipelajari Dengan Adaptive Neuro Fuzzy Inference System (ANFIS). Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07111750010011-Master_Thesis.pdf Download (6MB) | Preview |
Abstract
Tugas akhir ini membahas mengenai metode baru pendeteksian lokasi gangguan yang berbasis domain waktu. Dengan memberikan injeksi tegangan impulse pada sistem kelistrikan 20 Kv setiap detiknya di masing-masing fasa, maka akan diperoleh gambar respon sistem pada kondisi normal. Pada saat terjadi gangguan, maka respon tegangan impulse yang dinjeksikan akan berubah sehingga diperoleh gambar baru. Kemudian dilakukan perbandingan antara respon sebelum dan setelah gangguan, sehingga akan diketahui waktu dimana kedua respon tersebut berpisah satu dengan dengan yang lain, yang selanjutnya dijadikan dasar perhitungan untuk penentuan lokasi gangguan. Selanjutnya, hasil perbedaan waktu dan jarak ini akan dilakukan learning mengunakan Adaptive Neuro Fuzzy Inference System (ANFIS) untuk mendapatkan lokasi gangguan berada. Simulasi dilakukan menggunakan MATLAB 2016a, dengan beberapa kondisi diantarnya: memvariasi jenis gangguan (fasa ke tanah, fasa-fasa, dua fasa ke tanah, dan tiga fasa ke tanah), besar resistansi gangguan (5-300 ohm), resistansi ground (100-500 ohm), dan letak lokasi gangguan. Dari hasil simulasi diperoleh fakta bahwa nilai resistansi gangguan dan resistansi ground tidak mempengaruhi hasil deteksi lokasi gangguan dengan metode impulse. Penggunaan ANFIS menunjukkan perbaikan error lokasi, misalnya untuk gangguan 3 fasa ketanah pada jarak 1.853,35 km, error jarak mencapai 0.73% (13.49 m) yang sebelumnya errornya adalah 3.19% (59.12 m)
=======================================================================================================================================
This thesis described a new method for fault location detection in the time domain. This method works by injecting a train of impulse voltage in 20 kV electrical distribution system from the beginning of the feeder each phase in every
second via a coupling capacitor. The part of signal reflected by the network during normal conditions will be recorded by the device. When fault happened, there is a new discontinuity or mismatch impedance form the network, so the reflection response of the network will change as well. The algorithm is comparing the impulse response before and after fault to obtain time difference between the signals. The location of the fault can be calculated using a simple mathematic equation. In this thesis, the time difference is learned by an Adaptive Neuro-Fuzzy Inference System (ANFIS) to determine distance. The simulation will be done using MATLAB r2016a and will be used some conditions such as varying the type of the fault, value of fault resistance (5-300 Ω), ground resistance (100-500 Ω) and location of fault. Form the simulation result can be concluded that the value of fault resistance and fault resistance do not affect on the result of distance estimation which uses an impulse injection method. Application of ANFIS show that there is an improvement, for example: at distance 1.853,35 km, which is applied three phases has error distance about 0.73% (13,49 m). the error distance that use previous method has error about 3.19% (59.13 m).
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTE 621.319 Wid a-1 2020 |
| Uncontrolled Keywords: | Sistem Kelistrikan 20 kV, Deteksi lokasi gangguan, Injeksi tegangan impulse, Adaptive Neuro Fuzzy Inference System(ANFIS), |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK201 Electric Power Transmission |
| Divisions: | Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis |
| Depositing User: | Muhammad Budi Rahayu Widodo |
| Date Deposited: | 14 Mar 2025 04:17 |
| Last Modified: | 14 Mar 2025 04:17 |
| URI: | http://repository.its.ac.id/id/eprint/73433 |
Actions (login required)
 |
View Item |