A parameter estimation based sparse representation approach for mode separation and dispersion compensation of Lamb waves in isotropic plate

Abstract

A response Lamb wave is usually a combination of dispersive wave packets corresponding to different modes, and the explicit dispersion curves are hard to acquire accurately sometimes. A parameter estimation based sparse representation approach is established to extract each pure single-mode component with dispersion removal from the dispersive multi-mode Lamb waves. The dispersion characteristic parameters of the Lamb modes are extracted from a response Lamb wave with a known propagation distance using a parameter estimation algorithm. A parametric dictionary is designed to sparsely decompose the dispersive multi-mode Lamb waves, and those parameters used for constructing the dictionary are extracted from a response signal based on a Lamb wave propagation model. In the process of reconstructing each mode component, the original dictionary is replaced with a multi-mode and non-dispersive dictionary so that the dispersion effect can be removed. Results from simulations and experiments both verify the effectiveness of the proposed approach.

Accurate Characterization of the Adhesive Layer Thickness of Ceramic Bonding Structures Using Terahertz Time-Domain Spectroscopy

“…After the original echo signal is pre-processed by wavelet transform, the AR/MCM is used to further process the data according to Refs. [ 28 , 29 ]. The part of the signal with high signal-to-noise ratio is selected and the optimal order p is determined.…”

Weighted Structured Sparse Reconstruction-Based Lamb Wave Imaging Exploiting Multipath Edge Reflections in an Isotropic Plate

“…( ) g t  is the first-order time derivative of ( ) g t . The signal-to-noise ratio is increased through the above strategy of using the step-pulse excitation and post-processing technique [27]. The recovered narrowband responses to the tone burst centered at 80 kHz frequency are A0 mode dominant [28].…”

“…Narrowband responses to a five-cycle Hanning-windowed tone burst centered at 80 kHz are recovered from the original step-pulse responses using the post-processing technique [ 26 ] where is the desired narrowband response, is the Fourier transform, is the step-pulse response and is the first-order time derivative of . The signal-to-noise ratio is increased through the above strategy of using the step-pulse excitation and post-processing technique [ 27 ]. The recovered narrowband responses to the tone burst centered at 80 kHz frequency are A0 mode dominant [ 28 ].…”

Ultrasound Defect Localization in Shell Structures with Lamb Waves Using Spare Sensor Array and Orthogonal Matching Pursuit Decomposition

“…Furthermore, they furthered an interesting method to obtain the dispersion curve of an unknown material, in which they acquired the unit pulse response of the measured structure, and interpolated the transfer function from the frequency domain into the wavenumber domain [13]. After this, they systematically proposed the construction method of over-completed dictionaries, which contained dispersive atoms and nondispersive atoms [14].…”