Structured vector fitting framework for mechanical systems

In this paper, we develop a structure-preserving formulation of the data-driven vector fitting algorithm for case modally damped mechanical systems. Using structured pole-residue form transfer function second-order systems, propose two possible extensions barycentric formula system functions. Integrating these new forms within classical leads to algorithms that allow computation systems from data in least squares fashion. Thus, learned model is guaranteed have desired structure. We test proposed on benchmark models.