An adaptive extended Kalman filter for structural damage identification

The identification of structural damage is an important objective of health monitoring for civil infrastructures. System identification and damage detection based on measured vibration data have received intensive studies recently. Frequently, damage to a structure may be reflected by a change of some system parameters, such as a degradation of the stiffness. In this paper, we propose an adaptive tracking technique, based on the extended Kalman filter approach, to identify the structural parameters and their changes when vibration data involve damage events. The proposed technique is capable of tracking the changes of system parameters from which the event and severity of structural damage may be detected on-line. Our adaptive filtering technique is based on the current measured data to determine the parametric variation so that the residual error of the estimated parameters is contributed only by noise. This technique is applicable to linear and nonlinear structures. Simulation results for tracking the parametric changes of nonlinear elastic, hysteretic and linear benchmark structures are presented to demonstrate the application and effectiveness of the proposed technique in detecting structural damage, using measured vibration data.