Broadband computation of system output sensitivities using model-order reduction

This paper presents an efficient model-order reduction approach for computing fast frequency sweeps of system output sensitivities. The proposed framework allows the use of any projection-based order reduction method and renders the dimension of the reduced model independent of the number of parameters. Moreover, a thorough error analysis is presented which also clarifies the convergence properties of existing methods. A numerical experiment validates the authors´ findings.