Comparison of Three Time-domain Methods for Damage Tracking of Nonlinear Hysteretic Vibration Isolation System with Unknown Inputs

Abstract

Recently, vibration isolation technology has been widely applied in different types of structures, which refer to as nonlinear hysteretic vibration isolation systems, to withstand the strong vibration that may damage the structures, influent the structural normal operation, or reduce the life of structures. However, parameter identification and damage tracking of nonlinear hysteretic vibration isolation systems is quite challenging because the restoring behavior of these systems are inelastic. Besides, the external inputs of these systems may not be accurately measured or even may not be measurable in practical applications. To resolve the problems above, in this research, three timedomain analysis methods, which refer to as adaptive extended kalman filter with unknown inputs (AEKF-UI), adaptive sequential nonlinear least-square estimation with unknown inputs (ASNLSE-UI) and adaptive quadratic sum-square error with unknown inputs (AQSSE-UI), have been newly developed to online identify the structural parameters and to real-time track the structural damages. Further, experiment research was conduct through a series of laboratory tests using a scaled vibration-isolation structure model with two typical excitations in the case of different damage scenarios. Finally, the capability of AEKF-UI, ASNLSE-UI and AQSSE-UI methods in parameter identification and damage tracking have been demonstrated and compared in terms of accuracy, convergence and efficiency.