Sparse representation for Lamb-wave-based damage detection using a dictionary algorithm

Lamb waves are being investigated extensively for structural health monitoring (SHM) because of their characteristics of traveling long distances with little attenuation and sensitivity to minor local damage in structures. However, Lamb waves are dispersive, which results in the complex overlap of waveforms in the damage detection applications of the SHM community. This paper proposes a sparse representation strategy based on an l1-norm optimization algorithm for guided-Lamb-wave-based inspections. A comprehensive dictionary is designed containing various waveforms under diverse conditions so that the received waveform can be decomposed into a spatial domain for the identification of damage location. Furthermore, the l1-norm optimization algorithm is employed to pursue the sparse solution related to the physical damage location. The functionality of the created dictionary is validated by both metal beam and composite wind turbine experiments. The results indicate a great potential for the proposed sparse representation using a dictionary algorithm, which provides an effective alternative approach for damage detection.