Implementation of Support Vector Machine Algorithm in a Real-time BLDC Motor Bearing Fault Classification with Discrete Wavelet Transform as Feature Extractor

Brushless DC (BLDC) Motors are integral to industrial operations. Continuous motor usage can lead to various faults with significant consequences if left unaddressed. These faults may impact the motor, its surrounding system, disrupt economic activities, and potentially result in catastrophic failur...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2024 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET) S. 1 - 6
Hauptverfasser: Masangkay, Shan Joshua Raym C., Mendigoria, Lester Joseph B., Reyes, Ivan Ross M., Ostia, Conrado F.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 26.08.2024
Schlagworte:
Accuracy
Arduino Programming
Brushless DC Motor
Classification algorithms
Discrete Wavelet Transform
Discrete wavelet transforms
Fault diagnosis
Feature Classification
Feature extraction
Haar wavelet
MATLAB Simulink
Motor Fault Diagnosis
Motors
Real-time bearing fault diagnosis
Real-time systems
Support Vector Machine
Support vector machines
Testing
Wavelet analysis
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Brushless DC (BLDC) Motors are integral to industrial operations. Continuous motor usage can lead to various faults with significant consequences if left unaddressed. These faults may impact the motor, its surrounding system, disrupt economic activities, and potentially result in catastrophic failures. This study introduces a methodology combining Support Vector Machine (SVM) for feature classification with Discrete Wavelet Transform (DWT) for feature extraction. Through machine learning techniques, voltage signals from multiple BLDC motor samples with diverse faults were examined. Performance metrics, including precision, recall, accuracy, and F-1 scores, were calculated to evaluate the algorithm's effectiveness. The Support Vector Machine, trained alongside the Discrete Wavelet Transform, achieved an accuracy of 96.98 percent during validation and 90.37 percent during real-time testing. These results highlight the practical application of the proposed algorithm for efficient motor fault diagnosis.
DOI:10.1109/IICAIET62352.2024.10729943