Premium
Application of the unscented Kalman filter for real‐time nonlinear structural system identification
Author(s) -
Wu Meiliang,
Smyth Andrew W.
Publication year - 2007
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.186
Subject(s) - extended kalman filter , kalman filter , nonlinear system , control theory (sociology) , system identification , identification (biology) , invariant extended kalman filter , unscented transform , engineering , nonlinear filter , noise (video) , computer science , control engineering , filter (signal processing) , artificial intelligence , filter design , data modeling , computer vision , physics , botany , control (management) , quantum mechanics , biology , software engineering , image (mathematics)
Over the past few decades, structural system identification based on vibration measurements has attracted much attention in the structural dynamics field. The well‐known extended Kalman filter (EKF) is often used to deal with nonlinear system identification in many civil engineering applications. In spite of that, applying an EKF to highly nonlinear structural systems is not a trivial task, particularly those subject to severe loading. Unlike the EKF, a new technique, the unscented Kalman filter (UKF) is applicable to highly nonlinear systems. In this paper, the EKF and UKF are compared and applied for nonlinear structural system identification. Simulation results show that the UKF produces better state estimation and parameter identification than the EKF and is also more robust to measurement noise levels. Copyright © 2006 John Wiley & Sons, Ltd.