An efficient numerical simulation technique is introduced to extract the propagation characteristics of a millimeter guided wave structure. The method is based on the application of the Krylov subspace model order red...An efficient numerical simulation technique is introduced to extract the propagation characteristics of a millimeter guided wave structure. The method is based on the application of the Krylov subspace model order reduction technique (Padé via Lanczos) to the compact finite difference frequency domain (FDFD) method. This new technique speeds up the solution by decreasing the originally larger system matrix into one lower order system matrix. Numerical experiments from several millimeter guided wave structures demonstrate the efficiency and accuracy of this algorithm.展开更多
Reduction of complex protein networks models is of great importance.The accuracy of a passivity preserving algorithm (PRIMA) for model order reduction (MOR) is here tested on protein networks,introducing innovative va...Reduction of complex protein networks models is of great importance.The accuracy of a passivity preserving algorithm (PRIMA) for model order reduction (MOR) is here tested on protein networks,introducing innovative variations of the standard PRIMA method to fit the problem at hand.The reduction method does not require to solve the complete system,resulting in a promising tool for studying very large-scale models for which the full solution cannot be computed.The mathematical structure of the considered kinetic equations is preserved.Keeping constant the reduction factor,the approximation error is lower for larger systems.展开更多
文摘An efficient numerical simulation technique is introduced to extract the propagation characteristics of a millimeter guided wave structure. The method is based on the application of the Krylov subspace model order reduction technique (Padé via Lanczos) to the compact finite difference frequency domain (FDFD) method. This new technique speeds up the solution by decreasing the originally larger system matrix into one lower order system matrix. Numerical experiments from several millimeter guided wave structures demonstrate the efficiency and accuracy of this algorithm.
文摘Reduction of complex protein networks models is of great importance.The accuracy of a passivity preserving algorithm (PRIMA) for model order reduction (MOR) is here tested on protein networks,introducing innovative variations of the standard PRIMA method to fit the problem at hand.The reduction method does not require to solve the complete system,resulting in a promising tool for studying very large-scale models for which the full solution cannot be computed.The mathematical structure of the considered kinetic equations is preserved.Keeping constant the reduction factor,the approximation error is lower for larger systems.
