A Two Stage Identification Algorithm For Parametric Of Highly Nonlinear Structures With Incomplete Measurements

Ying Lei,Mingyu He,Lijun Liu
2014-01-01
Abstract:It is still challenging to identify highly nonlinear structures using partial measurements of structural responses. In this paper, a straightforward two- stage identification approach is proposed for this purpose. In the first stage, an equivalent linear system for a nonlinear system is established and identified using extended Kalman filter (EKF) with limited measurements. The locations of nonlinearity are identified by comparing the differences between the identified structural parameters of the equivalent linear system and those of the original linear system. In the second stage, the unscented Kalman filter (UKF) is used to identify the parameters of the highly nonlinear structural system. Several numerical simulation examples with different nonlinear models are used to validate the effectiveness of the proposed algorithm.
What problem does this paper attempt to address?