System identification and damage detection of structures

An important goal of any structural health monitoring system is to detect the structural damage in order to ensure the reliability and safety of the structure in addition to the minimization of the life-cycle cost. Time domain analysis techniques, including the least square estimation (LSE) and the extended Kalman filter (EKF), have been used for the identification of structural parameters. In practice, acceleration responses are usually measured on-line. Hence, the application of the LSE approach which may require the displacement measurements is not practical. With only the measurements of acceleration responses, the on-line system identification and damage detection is possible based on the EKF approach. However, in the EKF approach, the solutions (estimates) of the extended state vector after linearization of the highly nonlinear state equation may easily become unstable and may not converge. Likewise, the dimension of the extended state vector is quite large, and hence the computational efforts are quite involved.; In order to remove all the drawbacks of LSE and EKF approaches for the on-line system identification and damage detection, a new approach, referred to as the sequential nonlinear least square estimation (SNLSE), is proposed in this dissertation. An adaptive tracking technique recently proposed in the literature is implemented in the proposed SNLSE approach to track the structural damage. Further, an, extended SNLSE approach has been derived, referred to as the sequential nonlinear least square estimation with unknown inputs (excitations) and unknown outputs (responses) (SNLSE-UI-UO), to cover the general case, in which some external excitations and acceleration responses are not pleasured or not available. Simulation results using linear and nonlinear building models, as well as finite-element models are presented to demonstrate the accuracy of the proposed SNLSE and SNLSE-UI-UO in tracking the structural damage. Finally, the newly developed SNLSE-UI-UO method and the substructure approach are used to identify structural damages in complex structures, to demonstrate the feasibility of the local health monitoring for critical elements without the global information of the structure.; In this dissertation, another new approach, referred to as the quadratic sum squares error (QSSE), has been developed for the on-line system identification and damage detection. Further, the same adaptive tracking technique is implemented in the proposed QSSE approach to identify the time-varying system parameters of the structure, referred to as the adaptive quadratic sum squares error (AQSSE). This AQSSE approach has also been extended to cover the general case, in which the external excitations are not measured or not available, referred to as the adaptive quadratic sum squares error with unknown inputs (excitations) (AQSSE-UI). Simulation results for finite-element models of linear structures and nonlinear building models are presented to demonstrate the accuracy of the proposed AQSSE and AQSSE-UI in tracking the structural damage.