2025 Volume 71B Pages 57-65
In structural health monitoring, one key objective is to assess the integrity of buildings by observing their seismic response. Due to the limitations in the operation of monitoring systems, it is often impractical to monitor all stories, and the number of observed stories is typically limited. To predict the seismic response of unmonitored stories, we propose a method that translates the entire building into an equivalent shear model without relying on prior information on eigenmodes. The proposed method adjusts the top acceleration transfer function assumed to be derived by observed data to account for differences in higher-order natural circular frequencies between the frame model and the shear model. By fitting the transfer function, the stiffness and damping factors of the shear model are identified, enabling the estimation of maximum inter-story drifts and the absolute acceleration responses of the unmonitored stories. In a numerical example, the proposed system identification method is applied to a 20-story three-dimensional frame model, and the identification accuracy and effectiveness of the method are demonstrated.
