Romadhony, Khoirul (2023) Perancangan Sistem Deteksi Kesalahan Winding Short pada Electric Submersible Pump (ESP) menggunakan Machine Learning di Sumur Minyak UPB-12 PT. SIPL Gresik. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 02311940000148-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
02311940000148-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (3MB) | Request a copy |
Abstract
Penelitian ini menampilkan metode yang secara spesifik digunakan dalam mendeteksi kesalahan pada winding motor induk di electric submersible pump. Studi mengenai deteksi kesalahan pada pompa sudah sering dilakukan sebelumnya. Pendekatan atau metode yang digunakan pun beragam. Pemilihan dari metode diantara metode lainnya bergantung pada seberapa efektif, mudah digunakan dan efesiensinya. Pendeteksi yang dilakukan secara dini dalam menentukan fault adalah sebuah keunggulan. Pada penelitian ini mendeteksi kesalahan winding short motor induksi di electric submersible pump menggunakan machine learning untuk menentukan apakah terjadi kesalahan atau tidak. Parameter yang digunakan dalam mendeteksi kesalahan yaitu arus, temperatur winding, pressure input dan pressure discharge. Parameter ini nantinya akan digunakan dalam merancang pendeteksi untuk digunakan dalam mendeteksi data yang telah dikelompokkan menjadi data health, warning, dan fault yang nantinya akan diklasifikasikan menggunakan algoritma machine learning dengan algoritma SVM, ensamble, decision tree dan KNN. Algoritma ini nantinya akan dibandingkan dalam menentukan algoritma dengan akurasi terbaik dalam mengklasifikasikan kondisi motor induksi. Setelah diklasifikasikan data yang terklasifikasi sebagai warning akan dilakukan perhitungan remaining useful life untuk mengetahui lama waktu tersisa dari mesin hingga terjadinya fault.
=================================================================================================================================
This research presents a method that is specifically used to detect faults in the winding of the main motor in an electric submersible pump. Studies on fault detection in pumps have often been carried out before. The approaches or methods used also vary. The choice of a method over other methods depends on how effective, easy to use and efficient it is. Early detection in determining faults is an advantage. In this study, detecting short winding errors in induction motors in electric submersible pumps uses machine learning to determine whether an error has occurred or not. The parameters used in detecting faults are current, winding temperature, input pressure and discharge pressure. These parameters will later be used in designing detectors for use in detecting data that has been grouped into health, warning, and fault data which will later be classified using machine learning algorithms with SVM, ensemble, decision tree and KNN algorithms. This algorithm will later be compared in determining the algorithm with the best accuracy in classifying induction motor conditions. After classifying the data that is classified as a warning, the remaining useful life will be calculated to determine the remaining time of the machine until a fault occurs.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | electric submersible pump, machine learning, RUL, winding short motor induction |
| Subjects: | Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. T Technology > TJ Mechanical engineering and machinery > TJ174 Maintenance and repair of machinery T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Khoirul Romadhony |
| Date Deposited: | 05 Sep 2023 06:06 |
| Last Modified: | 05 Sep 2023 06:06 |
| URI: | http://repository.its.ac.id/id/eprint/101252 |
Actions (login required)
 |
View Item |