An improved stochastic subspace identification approach for automated operational modal analysis of high-rise buildings

Abstract

The Stochastic Subspace Identification (SSI) method has attracted significant attention in the field of Operational Modal Analysis (OMA) for large civil engineering structures such as high-rise buildings. State-space-based identification methods often encounter the challenge of spurious modes, leading to extensive efforts to mitigate this issue, commonly through stabilization diagram techniques. Nonetheless, the accuracy of the stabilization diagrams can be compromised by various interferences and the presence of spurious poles (i.e., outliers) in the high-order region. Consequently, this diminishes the precision of the modal identification. This study introduces an improved automated OMA technique based on SSI to enhance the reliability and accuracy of modal estimation. The proposed approach encompasses modal identification using SSI coupled with automated interpretation of SSI outputs, employing rigorous validation criteria and clustering techniques. An essential component of this approach involves signal segmentation during the data preprocessing stage, enabling the effective removal of spurious poles and thus a more robust interpretation of modal identification results via the stabilization diagram in an automated fashion. A numerical example was employed to demonstrate the effectiveness and applicability of the method. Subsequently, the proposed methodology is validated using field measurements conducted on two high-rise buildings.