Inference of the modal parameters of a walkway through the clustering and bootstrap techniques association

Resumo

Monitoring the behavior of a structure in operation is essential for the early detection of changes that
may compromise its integrity. This is possible from the identification of its modal parameters, estimated through
the modal analysis of the response signals (continuous in time) of the structure to the actions that act on it. Because
they are estimated, the parameters obtained have uncertainties that significantly interfere with the reliability of
structural monitoring, and therefore their quantification is essential. The principles of systems identification and
experimental estimation have provided, in the last decades, innovative tools for the understanding and control of
vibrations, design optimization, performance evaluation and structural integrity (Rainieri and Fabbrocino [1]).
Several studies have also ratified the efficiency of continuous monitoring of structures in assessing their integrity
(CHENG et al. [2] and SAISI et al. [3]). Thus, there is a need for study and continuous improvement of monitoring
and structural identification systems in order to ensure the safety of existing structures. The objective of this work
is to apply, in response signs of a pedestrian walkway submitted to people walking, a modal analysis methodology
that provides more robust estimates and with lower levels of uncertainty of its modal parameters, through the
association of Data- driven Stochastic Subspace Identification (SSI-DATA), bootstrap and clustering techniques.
Finally, the results found for the uncertainties of the estimated parameters are discussed.