Deteksi Kerusakan Bearing Pada Motor Induksi Tiga Fasa Menggunakan Metode EMD-ANN

Ewahyono, Bramantyo (2025) Deteksi Kerusakan Bearing Pada Motor Induksi Tiga Fasa Menggunakan Metode EMD-ANN. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022221017-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Penelitian ini menjelaskan deteksi kerusakan bearing (ball bearing, inner race, outer race) pada motor induksi tiga fasa menggunakan metode EMD-ANN. Kerusakan motor induksi dapat mengganggu proses produksi dan menimbulkan kerugian. Motor Current Signature Analysis (MCSA) merupakan metode yang umum digunakan untuk mendeteksi kerusakan pada motor listrik. Penelitian ini mengembangkan sistem berbasis MCSA dengan menganalisis sinyal menggunakan Empirical Mode Decomposition (EMD) dan Artificial Neural Network (ANN). EMD menguraikan arus motor menjadi Intrinsic Mode Functions (IMF), sedangkan Sample Between Zero Crossing (SBZC) dan Time Successive Between Zero Crossing (TSZC) digunakan untuk ekstraksi dan klasifikasi sinyal non-stasioner. ANN dikembangkan sebagai pengklasifikasi cerdas dengan input berupa parameter kurva Probability Density Function (PDF) dari simpangan baku titik persilangan nol IMF. Hasil menunjukkan akurasi 100% dengan Mean Squared Error (MSE) 0,00019698 saat dilatih dengan seluruh data, dan saat menggunakan 70% untuk pelatihan, 15% untuk validasi, dan 15% untuk pengujian mendapatkan hasil 82,5% dengan MSE 0,095618. Efisiensi ANN dipengaruhi oleh arsitektur dan kualitas data.
==================================================================================================================================
This study presents the detection of bearing faults (ball bearing, inner race, outer race) in three-phase induction motors using the EMD-ANN method. Faults in induction motors can disrupt production processes and cause losses. Motor Current Signature Analysis (MCSA) is a commonly used method for detecting faults in electric motors. This study develops an MCSA-based system by analyzing signals using Empirical Mode Decomposition (EMD) and Artificial Neural Network (ANN). EMD decomposes the motor current into Intrinsic Mode Functions (IMF), while Sample Between Zero Crossing (SBZC) and Time Successive Between Zero Crossing (TSZC) are used for extracting and classifying non-stationary signals. ANN is developed as an intelligent classifier with inputs in the form of Probability Density Function (PDF) curve parameters from the standard deviation of IMF zero-crossing points. The results show 100% accuracy with a Mean Squared Error (MSE) of 0.00019698 when trained with the entire dataset, and 82,5% accuracy with an MSE of 0.095618 when using 70% of the data for training, 15% for validation, and 15% for testing. ANN efficiency is influenced by the architecture and data quality.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Motor Induksi 3-Fasa, Motor Current Signature Analysis, Empirical Mode Decomposition, Artifical Neural Network, Probability Density Function Three-Phase Induction Motor, Motor Current Signature Analysis, Empirical Mode Decomposition, Artifical Neural Network, Probability Density Function.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Bramantyo Ewahyono
Date Deposited:	07 Aug 2025 03:07
Last Modified:	07 Aug 2025 03:07
URI:	http://repository.its.ac.id/id/eprint/127733

Actions (login required)

View Item