Detection of weld defects using ultrasonic-guided waves based on matching pursuit and density peak clustering

Abstract

Due to the fish scale surface of the weld seam, the guided wave dispersion is intensified, resulting in serious mode aliasing problem of the detected signal. It is difficult to analyze defect echo signal and locate defect accurately. To solve this problem, a new method of ultrasonic guided wave detection is proposed for weld defects based on matching pursuit and density peak clustering. First, according to the characteristics of the guided wave echo signal, a matching pursuit algorithm based on Morlet wavelet dictionary is established. The time-frequency analysis and parameter analysis of the obtained wavelet atoms are carried out to realize the modal separation and identification of the guided wave signal. Then, the similarity weight is introduced into the density peak algorithm to cluster the atoms obtained by sparse decomposition. The obtained clustering center is used to locate the weld defect. The validity of the method is proved by simulation and experiment. Finally, the experimental results show that the positioning error is 0.261% when the proposed method is used to detect the weld defects of 3-mm wide steel plate.