Frequency domain data merging in operational modal analysis based on least squares approach

Abstract

Assembling of multi-setup measurements emerges as a challenging problem in the structural health monitoring applications and may cause some important issues in the estimation of global modal parameters such as frequency, damping ratio and modal shape vector. To overcome this problem, a novel frequency domain pre-identification data merging method is proposed in this study. In the proposed methodology, to obtain a single measurement set, a least squares approach is employed resulting in a global response that is scaled from the multi-setup data. For the verification of the proposed merging procedure, one numerical, two experimental studies and one real data application have been conducted. The results obtained from the numerical, experimental and real data analysis indicate that the presented methodology provides rather high-quality estimations for multi-setup measurement problems. © 2020 Elsevier Ltd