Nonlinearity of dynamical systems poses significant challenges to conventional modal analysis techniques, necessitating alternative approaches for nonlinear model identification. We employed a methodology that leverages frequency response data to approximate the dynamics of arbitrary dynamical systems through the application of sparse regression techniques. This approach was applied to identify both a nonlinear system with a single degree of freedom and a linear system with multiple degrees of freedom. The results demonstrate that this approach can effectively yield a robust dynamic model without requiring prior knowledge of the underlying physical principles.
|
