Abstract (100 words or less): |  | This work presents an optimization-based procedure for modal identification of structures subjected to vibration. The problem is formulated as a nonlinear least square fit to minimize the error between the measured and calculated output variables. The proposed procedure integrates genetic and Nelder-Mead algorithms to locate the optimum modal parameters in the search domain. When a promising area is detected by the genetic algorithm, the Nelder-Mead algorithm starts with initial simplex in the localized domain. A methodology for the implementation of this algorithm is discussed in this paper. Two examples are used to illustrate the algorithm performance. It is shown that this methodology has a great advantage over most classical identification techniques. |