Bayesian structural identification of a hysteretic negative stiffness earthquake protection system using unscented Kalman filtering

Abstract

The application of Bayesian system identification in the context of a hysteretic negative stiffness system for seismic protection of structures is presented. The negative stiffness system employs the concept of apparent weakening to decrease the effective lateral stiffness of structures subjected to strong earthquakes, resulting in a significant reduction of the base shear and other seismic demands. In this paper, results from large-scale experimental testing performed by Rice University and the University at Buffalo - SUNY are used to estimate the parameters that define the nonlinear models of structures equipped with negative stiffness systems. For this purpose, an unscented Kalman filter for augmented-state nonlinear estimation is employed for structural identification. It is shown that the identified models have the capability to accurately predict the nonlinear hysteretic behavior of the modified structure. The predicted response quantities include lateral drifts and accelerations, base shear, restoring forces, and internal forces in structural members. The identified models provide benchmark parameters that can be used to predict the performance of negative stiffness systems, which is useful for the future design of structures equipped with this type of earthquake protection devices.