Version 1
: Received: 10 September 2024 / Approved: 10 September 2024 / Online: 10 September 2024 (11:07:18 CEST)
How to cite:
Kim, D.-Y.; Kareem, A. B.; Domingo, D.; HUR, J. A Comparative Data Augmentation-Assisted Diagnostic Framework for Industrial Centrifugal Pumps. Preprints2024, 2024090789. https://doi.org/10.20944/preprints202409.0789.v1
Kim, D.-Y.; Kareem, A. B.; Domingo, D.; HUR, J. A Comparative Data Augmentation-Assisted Diagnostic Framework for Industrial Centrifugal Pumps. Preprints 2024, 2024090789. https://doi.org/10.20944/preprints202409.0789.v1
Kim, D.-Y.; Kareem, A. B.; Domingo, D.; HUR, J. A Comparative Data Augmentation-Assisted Diagnostic Framework for Industrial Centrifugal Pumps. Preprints2024, 2024090789. https://doi.org/10.20944/preprints202409.0789.v1
APA Style
Kim, D. Y., Kareem, A. B., Domingo, D., & HUR, J. (2024). A Comparative Data Augmentation-Assisted Diagnostic Framework for Industrial Centrifugal Pumps. Preprints. https://doi.org/10.20944/preprints202409.0789.v1
Chicago/Turabian Style
Kim, D., Daryl Domingo and Jangwook HUR. 2024 "A Comparative Data Augmentation-Assisted Diagnostic Framework for Industrial Centrifugal Pumps" Preprints. https://doi.org/10.20944/preprints202409.0789.v1
Abstract
This study presents an advanced data augmentation framework to enhance diagnostics in centrifugal pumps using vibration data, with a focus on practical industrial applications. Initially, Gaussian noise and signal stretching were employed to generate synthetic data, addressing the common issue of insufficient defect data in industrial settings. These methods simulate varied operating conditions and different rotating speeds. Recognizing the limitations of traditional approaches, we further integrated advanced models—Long Short-Term Memory (LSTM), Autoencoder (AE), and Generative Adversarial Networks (GANs) to augment the dataset comprehensively. This enhanced approach increases the robustness and accuracy of machine-learning models for fault detection and classification. Experimental results show significant improvements in diagnostic algorithm performance, reducing false positives and increasing fault detection rates. The study offers a complete framework for incorporating both traditional and advanced data augmentation techniques into predictive maintenance protocols, ensuring reliable operation of centrifugal pumps in diverse industrial environments.
Keywords
centrifugal pump; data augmentation; electric motor; fault classification; gaussian noise; PHM
Subject
Engineering, Mechanical Engineering
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.