Parametric modelling of nonlinearities by covariance analysis

Abstract

An integrated approach for the identification of mechanical systems with nonlinear force clements is shown. The method is split in two parts. The first part based on the covariance method uses measured data from the system and a linear model to proof which nonlinearities have an influence on the dynamical behavior to get a nonlinear model. In the second part, the parameter of this nonlinear model will are identified using measured covariances of the system for different excitation levels.