Pipeline Flange Loose Bolt Localization Technology Based on Waveform Correlation Coefficient
Journal of Engineering System · 2026-06 · 4 (2) · 4-10
DOI: 10.62517/jes.202602202 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Engineering System, 2026-06. Verfasst von Cong Yu und 2 weiteren Autor:innen. Die Arbeit nennt 11 Literaturangaben.
Abstract (Originalsprache)
To address the challenge of locating bolt looseness in pipeline flanges equipped with sealing gaskets, this paper proposes a detection method based on stress wave principles combined with deep learning. Vibration signals were collected under various bolt loosening conditions at different excitation points on the flange. The overall waveform correlation coefficient and segmental waveform correlation coefficients of the signals were calculated to construct a feature dataset, and a Support Vector Machine (SVM) model was employed to identify the positions of loosened bolts. Experimental results demonstrate that when excitation points are selected near the sensor or far from the sensor, the waveform correlation coefficients across different time domains effectively reflect variations in the location of loosened bolts, achieving high recognition accuracy. This method provides a feasible technical approach for accurately locating bolt looseness in pipeline flanges.
Zitieren
APA
Yu, C., Cheng, P., & Shao, C. (2026). Pipeline Flange Loose Bolt Localization Technology Based on Waveform Correlation Coefficient. Journal of Engineering System, 4(2), 4-10. https://doi.org/10.62517/jes.202602202
Vancouver
Yu C, Cheng P, Shao C. Pipeline Flange Loose Bolt Localization Technology Based on Waveform Correlation Coefficient. Journal of Engineering System. 2026;4(2):4-10. doi:10.62517/jes.202602202
BibTeX
@article{Yu2026,
author = {Yu, Cong and Cheng, Peng and Shao, Chenxi},
title = {Pipeline Flange Loose Bolt Localization Technology Based on Waveform Correlation Coefficient},
journal = {Journal of Engineering System},
year = {2026},
volume = {4},
number = {2},
pages = {4--10},
doi = {10.62517/jes.202602202}
}
EUVIMED