Sahrin, Alfin (2017) Koordinasi Directional Overcurrent Rele Menggunakan Cascade Forward Neural Network Pada Jaring Distribusi Pembangkit Tersebar. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07111550010204-Mater_Thesis.pdf - Accepted Version Download (4MB) | Preview |
Abstract
Sistem proteksi dalam sistem jaring distribusi yang terhubung dengan pembangkit tersebar akan menimbulkan permasalahan baru yang perlu diselesaikan. Dengan menggunakan parameter arus sebagai dasar proteksi untuk setting overcurrent rele dengan sistem loop perlu dilengkapi elemen arah forward dan reverse sehingga menjadi directional overcurrent rele. Pada sistem jaring distribusi dengan pembangkit tersebar dimungkinkan akan terjadi perubahan topologi kombinasi jumlah pembangkit sehingga terjadi pula perubahan setting directional overcurrent rele. Solusi permasalahan tersebut dapat diselesaikan menggunakan sistem hirarki kontrol (hierarchical control) digunakan pengolahan, penginderaan dan penyesuaian, pemantauan dan pengawasan. Proses tersebut dapat dilakukan dengan cara melakukan pemodelan dengan menggunakan cascade forward neural network sehingga akan didapatkan hasil model jaringan saraf tiruan. Dari hasil model tersebut menghasilkan output nilai setting Ipickup dan TDS kemudian dilakukan pengujian dengan cara mensimulasikan pada software ETAP 12.6 dengan sistem IEEE 9 bus yang dimodifikasi dan di uji coba pada perangkat keras dengan sistem master dan slave. Proses pelatihan dan pengujian menggunakan metode lavenberg marquadt dengan mean square error 7,86e^(-14), dan data yang diproses pada master dapat dikirim ke slave dengan waktu rata – rata 31 detik.
==========================================================================================================
Protection systems in a distribution system connected to distributed generation will create new problems that need to be solved. Using current parameter as protection base for setting relay overcurrent with loop system need to have forward and reverse element then become directional overcurrent relay. In a distribution system with distributed generators it is possible to change the topology of the combined number of plants, so there is also a change of directional overcurrent relay settings. The solution of the problems can be solved using a hierarchical control system used for processing, sensing and adjusting, monitoring and monitoring. The process can be done by way of modeling by using cascade forward neural network so that it will produce artificial neural network model. From the results of the model produces output values of setting Ipickup and TDS then tested by simulating the ETAP 12.6 software with modified IEEE 9 bus system and tested on hardware with master and slave system. The training and testing process uses the lavenberg marquadt method with mean square error 7,86e^(-14), and data processed on the master can be sent to the slave with an average time of 31 seconds.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTE 621.317 Sah k-1 3100018074191 |
| Uncontrolled Keywords: | Koordinasi rele, directional overcuurent relay, jaringan saraf tiruan, kontrol hirarki; Relay coordination; directional overcurrent relay; artificial neural network; hierarchical control. |
| Subjects: | Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > T Technology (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2861 Electric relays |
| Divisions: | Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis |
| Depositing User: | Sahrin Alfin |
| Date Deposited: | 19 Feb 2018 04:17 |
| Last Modified: | 28 Apr 2020 05:09 |
| URI: | http://repository.its.ac.id/id/eprint/49288 |
Actions (login required)
 |
View Item |