Compressive sensing-based correlation sidelobe suppression for passive water pipeline fault detection using ambient noise. (15th July 2023)

• Record Type: Journal Article
• Authors: Li, Zhao; Lee, Pedro; Fink, Mathias; Murch, Ross; Davidson, Mark
• Is Part Of: Mechanical systems and signal processing. Volume 195(2023)
• Publication Date: 2023-07-15
• Subjects: Passive imaging -- Ambient noise -- Compressive sensing -- Correlation sidelobe -- Water pipelines -- Fault detection; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2023.110323
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: The ambient noise in water pipelines are observed as spontaneous signal sources that can be used for pipe fault detection by correlation analysis. However, the limited bandwidth of these noise signal causes strong correlation sidelobes, which introduces significant ambiguities when extracting the system response from correlation results and this increases the risk of false alarms from fault detections. This paper proposes a compressive sensing based method that can extend the noise bandwidth and suppress the correlation sidelobes. Numerical and field experiment results have shown that with the recovered wider bandwidth, the correlation sidelobes can be significantly suppressed and the pipe faults can be identified with greater certainty. The impacts of fault size as well as noise bandwidth, strength and spectrum features on the proposed method are also assessed through numerical experiments. Highlights: Compressive sensing method suppresses correlation sidelobes in the passive imaging. The method reduces the number of false peaks and the ambiguity in identifying faults. The impacts of noise bandwidth, SNR and PSD, and fault size are numerically assessed. The performance of the proposed method was verified by the field experiment.