Model-calibration-free damage identification of shear structures by measurement changes correction and sparse regularization

Abstract

This paper proposes a model-calibration-free method for damage identification of shear structures using the modal data. In doing so, a decoupling formulation is established upon the eigenparameter decomposition and to circumvent the ill-posedness, the sparse regularization is introduced. Due to the decoupling property, the sparse regularization parameter is directly determined through a simple threshold setting procedure. On the other hand, the model errors arise inevitably in acquiring the baseline model and they would give rise to unexpected errors in damage identification. To alleviate the effect of model errors, the measurement changes correction strategy is developed where the measured data used for damage identification is simply corrected by the pre-post measurement changes. It is remarkable that the model-calibration-free property of the proposed method is found just inherent from the measurement changes correction strategy because then there is no need to calibrate the mass and stiffness parameters of the structure and a rough uniform estimation of the mass and stiffness distributions is generally sufficient. Several well-established experiments are studied to verify the performance of the proposed method.