Predictive Maintenance Kerusakan Batang Rotor Pada Motor Induksi Tiga Fasa Dengan Metode Motor Current Signature Analysis Berbasis Artificial Neural Network

Sulistyo, Rayhan Noumi (2025) Predictive Maintenance Kerusakan Batang Rotor Pada Motor Induksi Tiga Fasa Dengan Metode Motor Current Signature Analysis Berbasis Artificial Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022211128-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Penelitian Motor induksi tiga fasa ini bertujuan untuk mengembangkan sistem predictive maintenance berbasis metode Motor Current Signature Analysis (MCSA) yang mengintegrasikan Fast Fourier Transform (FFT) dan Artificial Neural Network (ANN) untuk mendeteksi dan mendiagnosis kerusakan batang rotor pada motor induksi tiga fasa secara non-intrusif. MCSA digunakan untuk menganalisis harmonik arus stator, sementara FFT mentransformasi sinyal arus ke domain frekuensi untuk identifikasi kerusakan. Hasil data Arus Stator untuk nilai RMS telah didapatkan namun perlu pemrosesan data dari FFT untuk melihat frekuensi kerusakan batang rotor. Hasil pemrosesan FFT didapatkan dalam bentuk magnitude ampere kerusakan batang rotor di frekuensi 49.33Hz dan 50.67HZ, 46Hz dan 54Hz, serta 42.67Hz dan 57.33Hz. Algoritma ANN dirancang untuk memproses data sekuensial dan mendeteksi pola kerusakan secara akurat, bahkan dalam kondisi operasional yang dinamis. Sistem ini diuji melalui pengujian langsung pada variasi kondisi beban motor 0%, 50 %, dan 100%. Hasil penelitian melalui pembelajaran pola kerusakan dari ANN menunjukkan tingkat keakurasian pembacaan pola sebesar 95%
===================================================================================================================================
Research on three-phase induction motor that aims to develop a predictive maintenance system based on the Motor Current Signature Analysis (MCSA) method, integrating Fast Fourier Transform (FFT) and Artificial Neural Network (ANN) to detect and diagnose rotor bar faults in three-phase induction motors non-intrusively. MCSA is used to analyze the stator current harmonics, while FFT transforms the current signals into the frequency domain for fault identification. Stator current data in terms of RMS values have been obtained, but further FFT processing is needed to observe the fault frequencies of the rotor bars. The FFT processing results are presented as magnitude values (in amperes) of rotor bar faults at frequencies of 49.33Hz and 50.67Hz, 46Hz and 54Hz, as well as 42.67Hz and 57.33Hz. The ANN algorithm is designed to process sequential data and accurately detect fault patterns, even under dynamic operating conditions. The system was tested through direct experiments under varying motor load conditions from 0%, 50%, to 100% load. The research results, based on ANN's ability to learn fault patterns, show a pattern recognition accuracy rate of 95%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Motor Induksi Tiga Fasa, Motor Current Signature Analysis, pemeliharaan prediktif, Fast Fourier Transform, Artificial Neural Network Three Phase Induction Motor, Motor Current Signature Analysis, Predictive Maintenance, Fast Fourier Transform, Artificial Neural Network
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. 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:	Rayhan Noumi Sulistyo
Date Deposited:	25 Jul 2025 07:20
Last Modified:	25 Jul 2025 07:20
URI:	http://repository.its.ac.id/id/eprint/121681

Actions (login required)

View Item