Structure-preserving model order reduction based on Laguerre-SVD for coupled systems

摘要

In this paper, we present a model order reduction (MOR) method based on Laguerre polynomials and singular value decomposition (SVD) for coupled systems in the frequency domain. By constructing projection matrices from the global perspective and then blockdiagonalizing them, the reduced system is produced, which not only retains the structure of the original system, but also matches the first several Laguerre coefficients. In addition, the connection between our algorithm and the moment matching approximation is also discussed. The error estimation of our method is given as well. Besides, the stability of the reduced system is also studied. Finally, two numerical examples are provided to verify the effectiveness of our algorithm.