By Dourado, M.; Meireles, J.; Rocha, A.M.A.C.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
In this paper, the application of stochastic global optimization tech- niques, in particular the GlobalSearch and MultiStart solvers from MatLab®, to improve the updating of a structural dynamic model, are presented. For com- parative purposes, the efficiency of these global methods relatively to the local search method previously used in a Finite Element Model Updating program is evaluated. The obtained solutions showed that the GlobalSearch and MultiStart solvers are able to achieve a better solution than the local solver previously used, in the updating of a structural dynamic model. The results show also that the GlobalSearch solver is more efficient than the MultiStart, since requires less computational effort to obtain the global solution.