Seismic load identification for damper-added and base-isolated frame structures using EKF-UI algorithm and data fusion technique

Abstract

Timely acquisition of seismic loads is essential for post-earthquake safety assessments and disaster simulations of buildings. This paper proposes an alternative approach to obtain seismic waves by load identification method using structural responses instead of by ground motion sensors. Firstly, formulations are derived separately for both damper-added and base-isolated structures based on the extended Kalman filter algorithm, without prior information of structure or isolator parameters. In addition, displacement, velocity, and strain responses are incorporated to improve the algorithm noise robustness. Multi-rate data fusion is introduced to handle heterogeneous sensor sampling rates. Subsequently, numerical simulations for these two structures validate the feasibility of the proposed method and the contribution of data fusion technique on measurement noise resistance. Finally, a shaking table test performing on a five-floor frame with rubber isolation bearings further demonstrates the effectiveness and accuracy of the proposed identification algorithm for practical application. The focus on fusing supplementary data to mitigate noise for seismic load identification with incomplete measurements highlights the innovations.