ABSTRACT
An error or objective function can be defined as a weighted sum of the squared residual vector as
/ = A R r W A R (3)
where W is a weighting matrix that takes into account the different importance of some of the errors. Minimization of this function will maximize the agreement between the experimental and analytical model. The objective function can contain different criteria for finding out the discrepancies such as differences of frequencies, model coordinates, modal assurance criteria (MAC), coordinate model assurance criteria, frequency response assurance criteria, frequency domain assurance criteria and others (Wu 1999).
