ANALYSIS OF RSVD ALGORITHM USING HIGH FREQUENCY FRFs IN DAMAGE DETECTION

Abstract. The Robust Singular Value Decomposition algorithm (RSVD) was used to identify damaged structure using frequency response functions (FRFs). The experimental data were obtained from a laminate beam. The routine uses robust statistics and analysis was done in a data set from healthy and damaged structure. This technique has being developed for FRFs at high frequency range, which have showed to be more sensitive to small damages. In this work are presented results for detection algorithm in different frequency ranges, effectiveness of coherence signal applied to measurement dataset, FRF signal resolution, RSVD parameters and computational costs. Keywords: Singular value decomposition, Structural health monitoring, laminated structure, damage detection