A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data

This study analyzes the impact of different types of random noise applied in Denoising Autoencoder (DAE) training on fault diagnosis performance, with the aim of improving noise removal for vibration time series data. While conventional studies typically train DAEs using Gaussian random noise, such...

Full description

Saved in:
Bibliographic Details
Main Authors: Jun-gyo Jang, Soon-sup Lee, Se-Yun Hwang, Jae-chul Lee
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/12/6523
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849433421819936768
author Jun-gyo Jang
Soon-sup Lee
Se-Yun Hwang
Jae-chul Lee
author_facet Jun-gyo Jang
Soon-sup Lee
Se-Yun Hwang
Jae-chul Lee
author_sort Jun-gyo Jang
collection DOAJ
description This study analyzes the impact of different types of random noise applied in Denoising Autoencoder (DAE) training on fault diagnosis performance, with the aim of improving noise removal for vibration time series data. While conventional studies typically train DAEs using Gaussian random noise, such noise does not fully reflect the complex noise patterns observed in real-world industrial environments. Therefore, this study proposes a novel approach that uses high-frequency noise components extracted from actual vibration data as training noise for the DAE. Both Gaussian and high-frequency noise were used to train separate DAE models, and statistical features (mean, RMS, standard deviation, kurtosis, skewness) were extracted from the denoised signals. The fault diagnosis rates were calculated using One-Class Support Vector Machines (OC-SVM) for performance comparison. As a result, the model trained with high-frequency noise achieved a 0.0293 higher average F1-score than the Gaussian-based model. Notably, the fault detection accuracy using the kurtosis feature improved significantly from 26.22% to 99.5%. Furthermore, the proposed method outperformed the conventional denoising technique based on the Wavelet Transform, demonstrating superior noise reduction capability. These findings demonstrate that incorporating real high-frequency components from vibration data into the DAE training process is effective in enhancing both noise removal and fault diagnosis performance.
format Article
id doaj-art-37ebaef54df8464f921775e119f3fbf1
institution Kabale University
issn 2076-3417
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-37ebaef54df8464f921775e119f3fbf12025-08-20T03:27:02ZengMDPI AGApplied Sciences2076-34172025-06-011512652310.3390/app15126523A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series DataJun-gyo Jang0Soon-sup Lee1Se-Yun Hwang2Jae-chul Lee3ADIALAB, 702-1, 57 Centum-dong-ro, Haeundae-gu, Busan 48059, Republic of KoreaDepartment of Naval Architecture and Ocean Engineering, College of Ocean Sciences, Gyeongsang National University, 11-dong, 2 Tongyeonghaean-ro, Tongyeong-si 53064, Republic of KoreaDepartment of Naval Architecture and Ocean Engineering, College of Ocean Sciences, Gyeongsang National University, 11-dong, 2 Tongyeonghaean-ro, Tongyeong-si 53064, Republic of KoreaDepartment of Naval Architecture and Ocean Engineering, College of Ocean Sciences, Gyeongsang National University, 11-dong, 2 Tongyeonghaean-ro, Tongyeong-si 53064, Republic of KoreaThis study analyzes the impact of different types of random noise applied in Denoising Autoencoder (DAE) training on fault diagnosis performance, with the aim of improving noise removal for vibration time series data. While conventional studies typically train DAEs using Gaussian random noise, such noise does not fully reflect the complex noise patterns observed in real-world industrial environments. Therefore, this study proposes a novel approach that uses high-frequency noise components extracted from actual vibration data as training noise for the DAE. Both Gaussian and high-frequency noise were used to train separate DAE models, and statistical features (mean, RMS, standard deviation, kurtosis, skewness) were extracted from the denoised signals. The fault diagnosis rates were calculated using One-Class Support Vector Machines (OC-SVM) for performance comparison. As a result, the model trained with high-frequency noise achieved a 0.0293 higher average F1-score than the Gaussian-based model. Notably, the fault detection accuracy using the kurtosis feature improved significantly from 26.22% to 99.5%. Furthermore, the proposed method outperformed the conventional denoising technique based on the Wavelet Transform, demonstrating superior noise reduction capability. These findings demonstrate that incorporating real high-frequency components from vibration data into the DAE training process is effective in enhancing both noise removal and fault diagnosis performance.https://www.mdpi.com/2076-3417/15/12/6523Denoising Autoencoderfault diagnosisvibration signalnoise filteringOne-Class Support Vector Machine
spellingShingle Jun-gyo Jang
Soon-sup Lee
Se-Yun Hwang
Jae-chul Lee
A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
Applied Sciences
Denoising Autoencoder
fault diagnosis
vibration signal
noise filtering
One-Class Support Vector Machine
title A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
title_full A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
title_fullStr A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
title_full_unstemmed A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
title_short A Study on Denoising Autoencoder Noise Selection for Improving the Fault Diagnosis Rate of Vibration Time Series Data
title_sort study on denoising autoencoder noise selection for improving the fault diagnosis rate of vibration time series data
topic Denoising Autoencoder
fault diagnosis
vibration signal
noise filtering
One-Class Support Vector Machine
url https://www.mdpi.com/2076-3417/15/12/6523
work_keys_str_mv AT jungyojang astudyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT soonsuplee astudyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT seyunhwang astudyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT jaechullee astudyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT jungyojang studyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT soonsuplee studyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT seyunhwang studyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata
AT jaechullee studyondenoisingautoencodernoiseselectionforimprovingthefaultdiagnosisrateofvibrationtimeseriesdata