Abstract:Aiming at the problems of strong manual dependence and low degree of automation in the finite element model correction of MIDAS Civil, a parameter dynamic calibration framework based on the MIDAS Civil API and Kriging surrogate model is proposed to achieve the standardization and full-process automation of model correction. By utilizing a Python-driven API to directly invoke the original design model, on the basis of maintaining the topological relationship unchanged, the batch iterative updates of parameters and automatic parsing of response data are achieved to avoid the inefficiency of traditional manual correction and the complexity of cross-platform modeling. Taking the Qinghai North Konggang Road No.2 Bridge as an engineering case, the static displacement error of the model after correction is reduced to 0.45% - 2.43%, and the frequency error is converged to 1.98%~3.87%. The corrected model can be directly reused in design optimization, health monitoring and other full lifecycle stages, providing an inheritable technical pathway for constructing the digital twin models of bridges.
