Experimental investigation of a Kriging surrogate-based finite element model updating method for bridge structures

Abstract

Finite element (FE) model updating methods are frequently used to evaluate the performance of a bridge. A precise initial analytical model is necessary to obtain accurate results in conventional FE model updating methods. However, a precise model consumes excessive calculation time. In this study, a new FE model updating method based on a surrogate model is presented to significantly reduce the calculation time without sacrificing accuracy. A Kriging surrogate model (or response surface model) is selected as a result of considering various kinds of surrogate models. In order to efficiently construct an accurate Kriging surrogate model, the sequential sampling strategy is introduced. The performance and efficiency of the proposed method are verified through a lab-scale experiment with a small-scale bridge model. It is demonstrated that the proposed FE model updating method has much better performance in the calculation time than the conventional one.