Software and Data for Structural Model Updating
Field dynamic testing on an as-built structure usually provides modal properties that are different from these generated by a finite element (FE) model. To update the FE model parameters, optimization problems can be formulated to minimize the difference between experimental and simulated modal properties. Various FE model parameters can be selected as optimization variables for model updating, such as the elastic moduli of structural members, as well as stiffness values of support springs.
This software package for FE model updating offers two modal property difference formulations, one using MAC values and the other using direct differences between eigenvectors. For each updating formulation, analytical Jacobian derivative of the objective function is implemented. To find the optimal solution of the formulated optimization problem, the package supports two optimization algorithms implemented in MATLAB lsqnonlin optimization toolbox, namely the Levenberg-Marquardt and the trust-region-reflective algorithms. Randomly generated starting values of the optimization variables can be adopted to increase the chance of finding global minimum of the optimization problem. Finally, the MATLAB code contains two structural examples to evaluate the model updating formulations and optimization algorithms.
The MATLAB source code and data are available for download at https://github.com/ywang-structures/Structural-Model-Updating.git.
Companion Monograph: Dong, X. and Wang, Y. (2018). "Formulation and optimization algorithm comparison for the FE model updating of large-scale structures."