Parameter estimation of structural dynamic models using eigenvalue and eigenvector information.

摘要

Structural system identification methods are analytical techniques for reconciling test data with analytical models. The response data frequently used to compare a finite element model and test data are the eigenvalues of the system. However, eigenvalues alone cannot assure an adequate model. Eigenvectors also provide valuable information for the process of updating finite element models. For large order, complex finite element models, ad-hoc procedures have often proven inadequate for model parameter updating. Therefore, parameter estimation techniques such as Bayes estimation or mathematical programming have been applied. Mathematical programming techniques can be use for parameter estimation allowing a very general definition of the objective function and constraints. This paper will present the application of mathematical programming techniques of parameter estimation to the updating of a finite element model of an electronic package. The following topics will be discussed in the paper. The mathematical programming formulation of the parameter estimation problem, which uses both eigenvalue and eigenvector response data. The software implementation of this technique. The application of this methodology to the estimation of parameters of an electronics package model.