Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process
The increased demand for predictive monitoring in Laser Beam Welding (LBW) process for quality management of the welded joint has led to the development of systems utilizing various signals with the combination of acoustic and optical signals being particularly cost-effective and less complex. Howev...
保存先:
| 主要な著者: | , , , , |
|---|---|
| フォーマット: | 論文 |
| 言語: | English |
| 出版事項: |
Elsevier Ltd
2025
|
| 主題: | |
| オンライン･アクセス: | http://umpir.ump.edu.my/id/eprint/43613/1/Enhancement%20of%20weld%20penetrations%20classification%20model%20performance.pdf http://umpir.ump.edu.my/id/eprint/43613/ https://doi.org/10.1016/j.apacoust.2024.110503 https://doi.org/10.1016/j.apacoust.2024.110503 |
| タグ: |
タグ追加
タグなし, このレコードへの初めてのタグを付けませんか!
|
| id |
my.ump.umpir.43613 |
|---|---|
| record_format |
eprints |
| spelling |
my.ump.umpir.436132025-01-17T03:53:40Z http://umpir.ump.edu.my/id/eprint/43613/ Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process Aleem, S.A.A. Mohd Fadhlan, Mohd Yusof M., Ishak F. R. M., Romlay I., Ishak TJ Mechanical engineering and machinery TL Motor vehicles. Aeronautics. Astronautics The increased demand for predictive monitoring in Laser Beam Welding (LBW) process for quality management of the welded joint has led to the development of systems utilizing various signals with the combination of acoustic and optical signals being particularly cost-effective and less complex. However, in a harsh production circumstance, both methods suffer from the noise problem, which could affect the effectiveness of the prediction model. This project aims to enhance a classification model to estimate the weld penetration conditions of LBW by employing the EMD-Z-score (EMD-ZS) and spectral averaging (SA) denoising algorithm in the pre-processing stage. To achieve the aims of this research, an experiment has been conducted through the manipulation of laser power and weld speed to obtain several penetration conditions such as full penetration, half penetration, incomplete penetration, and overheat penetration. The acoustic signal and optical spectrum were acquired during the process. The study found that the acoustic average amplitude and optical spectrum intensity are linearly proportional to penetration depth. Due to the process instability, a significant amount of spatters appear and exhibit a non-stationary random noise in both signals which leads to the high discrepancies of the extracted features value. This caused difficulties in distinguishing different weld penetration groups from their trend. The result of implementing EMD-ZS and SA showed that out of the 15 features analysed, 12 showed improvement after denoising, with the biggest improvement recorded to be 68.17%. The feature selection analysis identified Intensity and Log Different Absolute Standard Deviation (LDASD) as the most significant features in the original signal. In contrast, Intensity and Absolute Value of Summation of Square Root (AVSSR) were identified as the most significant features in the denoised signals. When comparing the original model to the denoised, the average accuracy improved from 97.4% to 99.8% in the training set and from 95.5% to 99.0% in the test set. Additionally, the precision improved by 53.3% in the training set and 11.17% for the test set, which indicates greater consistency and reliability in the denoised model. This research is expected to advance the LBW process by providing a predictive monitoring system to enhance the quality and efficiency with potential benefits for defect detection in various industries, ultimately leading to cost savings and improved product quality. Elsevier Ltd 2025-03 Article PeerReviewed pdf en http://umpir.ump.edu.my/id/eprint/43613/1/Enhancement%20of%20weld%20penetrations%20classification%20model%20performance.pdf Aleem, S.A.A. and Mohd Fadhlan, Mohd Yusof and M., Ishak and F. R. M., Romlay and I., Ishak (2025) Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process. Applied Acoustics, 231 (110503). pp. 1-18. ISSN 0003-682X. (In Press / Online First) (In Press / Online First) https://doi.org/10.1016/j.apacoust.2024.110503 https://doi.org/10.1016/j.apacoust.2024.110503 |
| institution |
Universiti Malaysia Pahang Al-Sultan Abdullah |
| building |
UMPSA Library |
| collection |
Institutional Repository |
| continent |
Asia |
| country |
Malaysia |
| content_provider |
Universiti Malaysia Pahang Al-Sultan Abdullah |
| content_source |
UMPSA Institutional Repository |
| url_provider |
http://umpir.ump.edu.my/ |
| language |
English |
| topic |
TJ Mechanical engineering and machinery TL Motor vehicles. Aeronautics. Astronautics |
| spellingShingle |
TJ Mechanical engineering and machinery TL Motor vehicles. Aeronautics. Astronautics Aleem, S.A.A. Mohd Fadhlan, Mohd Yusof M., Ishak F. R. M., Romlay I., Ishak Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| description |
The increased demand for predictive monitoring in Laser Beam Welding (LBW) process for quality management of the welded joint has led to the development of systems utilizing various signals with the combination of acoustic and optical signals being particularly cost-effective and less complex. However, in a harsh production circumstance, both methods suffer from the noise problem, which could affect the effectiveness of the prediction model. This project aims to enhance a classification model to estimate the weld penetration conditions of LBW by employing the EMD-Z-score (EMD-ZS) and spectral averaging (SA) denoising algorithm in the pre-processing stage. To achieve the aims of this research, an experiment has been conducted through the manipulation of laser power and weld speed to obtain several penetration conditions such as full penetration, half penetration, incomplete penetration, and overheat penetration. The acoustic signal and optical spectrum were acquired during the process. The study found that the acoustic average amplitude and optical spectrum intensity are linearly proportional to penetration depth. Due to the process instability, a significant amount of spatters appear and exhibit a non-stationary random noise in both signals which leads to the high discrepancies of the extracted features value. This caused difficulties in distinguishing different weld penetration groups from their trend. The result of implementing EMD-ZS and SA showed that out of the 15 features analysed, 12 showed improvement after denoising, with the biggest improvement recorded to be 68.17%. The feature selection analysis identified Intensity and Log Different Absolute Standard Deviation (LDASD) as the most significant features in the original signal. In contrast, Intensity and Absolute Value of Summation of Square Root (AVSSR) were identified as the most significant features in the denoised signals. When comparing the original model to the denoised, the average accuracy improved from 97.4% to 99.8% in the training set and from 95.5% to 99.0% in the test set. Additionally, the precision improved by 53.3% in the training set and 11.17% for the test set, which indicates greater consistency and reliability in the denoised model. This research is expected to advance the LBW process by providing a predictive monitoring system to enhance the quality and efficiency with potential benefits for defect detection in various industries, ultimately leading to cost savings and improved product quality. |
| format |
Article |
| author |
Aleem, S.A.A. Mohd Fadhlan, Mohd Yusof M., Ishak F. R. M., Romlay I., Ishak |
| author_facet |
Aleem, S.A.A. Mohd Fadhlan, Mohd Yusof M., Ishak F. R. M., Romlay I., Ishak |
| author_sort |
Aleem, S.A.A. |
| title |
Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| title_short |
Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| title_full |
Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| title_fullStr |
Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| title_full_unstemmed |
Enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| title_sort |
enhancement of weld penetrations classification model performance through the implementation of denoising methods towards the acquired acoustic and optical signals during laser beam welding process |
| publisher |
Elsevier Ltd |
| publishDate |
2025 |
| url |
http://umpir.ump.edu.my/id/eprint/43613/1/Enhancement%20of%20weld%20penetrations%20classification%20model%20performance.pdf http://umpir.ump.edu.my/id/eprint/43613/ https://doi.org/10.1016/j.apacoust.2024.110503 https://doi.org/10.1016/j.apacoust.2024.110503 |
| _version_ |
1827518376793079808 |
| score |
13.251813 |