Application of Compressed Sensing in Damage Imaging by Guided Wavefield in Metal Plate

As a non-contact method for the extraction of vibration signals of guided waves with high spatial resolution, Scanning Laser Doppler Vibrometer (SLDV) is very attractive for damage detection in guided wave. Conventional full wavefield detection methods of SLDV often require scanning point by point, which limit the efficiency of detection. By introducing compressed sensing method, an ultrasonic wavefield reconstruction technique is developed to realize damage localization and imaging in a metal plate. By developing the sparse characteristics of the wavefield, the discrete samples of the wavefield are obtained by random sampling under the condition of much smaller than the Nyquist sampling ratio, and then the wavefield is reconstructed by a nonlinear reconstruction algorithm. Coherence coefficient of power spectral density is established to evaluate the wavefield reconstruction effect. Experiment results show that hole-edge crack can be evaluated well in aluminum plate.