Rancang Bangun Sistem Deteksi Kesalahan Angular Misalignment Pada Motor Induksi Dengan Deep Learning Berdasarkan Data Pengukuran Arus

Afiifah, Adhela (2022) Rancang Bangun Sistem Deteksi Kesalahan Angular Misalignment Pada Motor Induksi Dengan Deep Learning Berdasarkan Data Pengukuran Arus. Other thesis, Institut Teknologi Sepuluh Nopember Surabaya.

Text 02311840000052-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 September 2024. Download (2MB) | Request a copy

Abstract

Motor induksi dianggap relatif handal dan kuat. Akan tetapi, keandalan perangkat elektromekanis ini masih rentan terhadap banyaknya jenis gangguan. Salah satunya adalah gangguan yang menyebabkan timbulnya misalignment. Upaya mitigasi risiko wajib dilakukan untuk mendeteksi dan mengelola risiko yang mungkin akan terjadi dalam penggunaan motor induksi. Pada penelitian ini dilakukan perancangan sistem deteksi kesalahan angular misalignment dengan deep learning pada kondisi sudut sebesar 1⁰, 2⁰ dan 3⁰. Variabel pengukuran yang digunakan adalah arus stator D1 pada motor induksi, mengacu pada Motor Current Signature Analysis. Pengukuran data arus dibaca oleh modul sensor SCT kemudian diolah menjadi Recurrence Plotting sebagai input algoritma deep learning. Arsitektur deep learning yang digunakan adalah 2 Dimension Deep Convolutional Neural Networks. Sistem deteksi kesalahan ini mampu mengklasifikasikan kondisi 2° dan 3° secara akurat. Hasil yang didapatkan dari klasifikasi sistem deteksi kesalahan memiliki tingkat akurasi sebesar 95,1% dengan nilai sensitifitas dan spesifisitas pada sistem deteksi ini berturut-turut yaitu 0,95 dan 0,98
=======================================================================================================================================
Induction motors are relatively reliable and powerful. However, the reliability of this electromechanical device is still vulnerable to many types of interference. One of them is a disturbance that causes misalignment. Risk mitigation efforts must be carried out to detect and manage risks that may occur in the use of induction motors. In this research, an angular misalignment fault detection system was designed with deep learning at angle variations of 1⁰, 2⁰ and 3⁰. The measurement variable used is the stator current D1 on the induction motor, referring to the Motor Current Signature Analysis. The current measurement data is obtained by SCT sensor module and then being processed into Recurrence Plotting as an input for deep learning algorithm. The deep learning architecture used is 2 Dimension Deep Convolutional Neural Networks. This fault detection system is able to accurately classify 2° and 3° conditions. The results obtained from the classification of the error detection system have an accuracy rate of 95.1% with the sensitivity and specificity values of the detection system being 0.95 and 0.98, respectively

Item Type:	Thesis (Other)
Additional Information:	621.381 536 Afi r-1
Uncontrolled Keywords:	Angular misalignment, Deep Learning, Motor Induksi
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	EKO BUDI RAHARJO
Date Deposited:	08 Dec 2022 01:40
Last Modified:	08 Dec 2022 01:40
URI:	http://repository.its.ac.id/id/eprint/95185

Actions (login required)

View Item