Enhancing the performance of stochastic subspace identification method via energy-oriented categorization of modal components

As an advanced modal identification method, Stochastic Subspace Identification (SSI) technique has been widely used in the engineering field in recent years. However, to apply this technique properly, the characteristics of external force acting on the structural system and the data quality of response signals should meet certain requirements, which sometimes cannot be well satisfied. This may result in some problematic issues, such as mode absences. To improve the working performance of standard SSI method, this study proposes an implementation strategy which is characterized by categorization of modal components in the response into different energy groups for data preprocessing. A numerical case study is presented first, to demonstrate that a key reason for the problematic issues encountered by the standard SSI technique is the imbalanced distribution of energies associated with different modal components of the system response. The proposed method is then detailed. To implement this method, the system response should be first divided into a number of component groups with each group consisting of certain modal components, whose spectral energies are in similar order of magnitude. Then, the standard SSI technique is utilized to identify the modal parameters for each of the above component groups. Measured responses from a real super-tall building are used in the proposed method and other modal identification techniques. The results are compared to validate and evaluate the performance of the proposed method.

Automated summary

Stochastic Subspace Identification (SSI) technique, an advanced modal identification method widely used in engineering, requires certain requirements to be met. This study proposes an implementation strategy to improve the working performance of the standard SSI method by categorizing modal components into different energy groups for data preprocessing.