Guided-wave signal processing using chirplet matching pursuits and mode correlation for structural health monitoring

Signal processing algorithms for guided wave pulse echo-based SHM must be capable of isolating individual reflections from defects in the structure, if any, which could be overlapping and multimodal. In addition, they should be able to estimate the time-frequency centers, the modes and individual energies of the reflections, which would be used to locate and characterize defects. Finally, they should be computationally efficient and amenable to automated processing. This work addresses these issues with a new algorithm employing chirplet matching pursuits followed by a mode correlation check for single point sensors. Its theoretical advantages over conventional time-frequency representations in all aspects are elaborated and these are demonstrated using numerical simulations and experiments in isotropic plate structures. The issue of in-plane triangulation is then discussed and experimental work done to explore this issue is presented. This work concludes with a description of how the algorithm can be extended to composite plate structures.