Core Concepts
The author proposes a method combining companion linearization and tensor decomposition to construct a reduced order model for parameterized linear systems, offering an efficient way to generate snapshots and accurate models.
Abstract
The content discusses approximating solutions to parameterized linear systems using Chebyshev HOPGD with sparse grid sampling. It combines companion linearization with the Krylov subspace method to construct a reduced order model. The method is detailed through numerical examples of a parameterized Helmholtz equation, showcasing competitive results.
The work focuses on generating snapshot solutions efficiently by decomposing a tensor matrix of precomputed solutions. It highlights the importance of selecting nodes carefully to ensure convergence and accuracy in the model construction process. The proposed approach offers flexibility in generating snapshots and constructing accurate models for parameterized linear systems.
Key points include the use of Chebyshev HOPGD with sparse grid sampling, combining companion linearization and tensor decomposition, numerical simulations of a parameterized Helmholtz equation, and the significance of node selection in model construction.
The content emphasizes the efficiency and robustness of the proposed method in generating accurate models for parameterized linear systems through careful selection of nodes during snapshot generation.
Stats
Specifically, the system arises from a discretization of a partial differential equation.
Numerical examples show competitiveness in solving a parameter estimation problem.
All experiments were carried out on a 2.3 GHz Dual-Core Intel Core i5 processor with 16 GB RAM.
Quotes
"The simulations are reproducible, and the software is available online."