Modified Quadratic Compression Method for mass and stiffness updating

The paper presents the modified Quadratic Compression Method (QCM) for both mass and stiffness model updating. The modeling error is defined in a parametric setup, i.e. with pre-specified principal submatrices multiplied by unknown scalar parameters. The optimal parameters are obtained by minimizing the error in a squared down version of the eigenvalue equation, and of the mass orthogonality condition, thus with reduced computation yet with no loss of information. The method has closed ties with Minimization of the Error in the Constitutive Equation (MECE), and in some cases is shown to belong to that class with a particular choice of the weighting matrix. Theoretical analysis of the propagation of the noise into the identified parameters, as well as extensive simulations, reveal that QCM has in some cases desirable noise filtering properties.