A defect localization method based on self-sensing and orthogonal matching pursuit

Yuqing Gao; Weilei Mu; Fuh-Gwo Yuan; Guijie Liu

doi:10.1016/j.ultras.2022.106889

A defect localization method based on self-sensing and orthogonal matching pursuit

Ultrasonics. 2023 Feb:128:106889. doi: 10.1016/j.ultras.2022.106889. Epub 2022 Nov 11.

Authors

Yuqing Gao¹, Weilei Mu², Fuh-Gwo Yuan³, Guijie Liu¹

Affiliations

¹ College of Engineering, Ocean University of China, Qingdao, Shandong 266100, China.
² College of Engineering, Ocean University of China, Qingdao, Shandong 266100, China; Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC 27695, USA. Electronic address: muweilei@ouc.edu.cn.
³ Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC 27695, USA.

PMID: 36395537
DOI: 10.1016/j.ultras.2022.106889

Abstract

In conventional structural health monitoring (SHM), a sensor array enables to localize a potential defect by using at least three lead zirconate titanate (PZT) patches. To reduce the vast number of patches needed for large-scaled structure, this paper presents an extremely sparse sensor array with only one single PZT patch, which could actuate and sense simultaneously. Firstly, a half-bridge circuit, referred as a self-sensing circuit is developed with a capacitor connected with the PZT patch, and the capacitance parameter and self-sensing performance are studied subsequently. Then, an orthogonal matching pursuit (OMP)-based sparse decomposition and dispersion removal algorithm is proposed to separate and reconstruct wave packets which are acutely overlapped. Subsequently, a matching strategy is proposed to determine the matching relationship between wave packets and wave paths. Finally, the ellipse-type imaging approach is employed to image the defect location. Two cases: one and two defects respectively are implemented to verify its efficacy. Experimental results illustrate that the proposed self-sensing unit and signal process method could erase the adverse effect of sensor-actuator interval and dispersion characteristic to the localization resolution and accuracy.

Keywords: Defect localization; Self-sensing; Wave packets decomposition.