Sensitivity of damage identification for multi-story structures by modified EKF method

Document Type

Conference Proceeding

Publication Date

1-1-2014

Abstract

In this study, the extended Kalman filter (EKF) is modified by using the derivative-free state estimation method, the weighted global iterative method and the fuzzy inference system to overcome practical problems encountered in structural identification. The modified extended EKF technique is then use to track the inter-story stiffness of buildings. The equivalent two-dimensional shear-type model and the decoupled equation of motion are adopted to simplify the procedure of structural identification. The effects of structural damping, noise of seismic records and location of limited seismic sensors on the identified inter-story stiffness are analyzed. It can be noted that the damping ratio has a slight effect on the accuracy of damage identification. The noise level in seismic records has a remarkable influence on the accuracy of identified inter-story stiffness. Arrangement of sensor locations can cause the significant identification errors of interstory stiffness in the condition of incomplete data system, particular at the floors having a great change of inter-story stiffness compared with adjacent floors. The effectiveness of modified EKF method is also verified by stiffness identification analysis of a shear-bending type building. The results show that EKF method combined with shear-type model may use as an effective tool to assess possible damage of the shear-bending type multistory buildings.

Publication Title

NCEE 2014 - 10th U.S. National Conference on Earthquake Engineering: Frontiers of Earthquake Engineering

Share

COinS