Compact Rational Krylov for Parametrized Systems with Application to BEM Frequency Sweeping

2026-07-08

AI Summary

This work addresses the computational challenge of repeatedly solving parameter-dependent linear systems, such as those arising in multi-frequency Helmholtz scattering problems within the boundary element method. The authors propose an efficient joint solution strategy that linearizes the original problem into a sequence of shifted linear systems and solves them using a rational Krylov GMRES algorithm with left–right preconditioning within the CORK framework. This study extends the CORK approach for the first time to parametrized linear systems, accommodating parameter-dependent right-hand sides, adaptive shift selection, and inexact inner solves, while exploiting the data-sparsity structure inherent in boundary element matrices. Numerical experiments demonstrate that the proposed method substantially reduces the computational cost of frequency sweeps, achieving high accuracy and excellent scalability.

Abstract

In parametrized linear systems $\mathsf{P}(μ)\mathsf{x}=\mathsf{b}$ the system matrix $\mathsf{P}$ depends nonlinearly on a parameter $μ$ and solutions are sought for many values of this parameter. We show that the compact rational Krylov (CORK) framework, originally introduced to solve nonlinear eigenvalue problems, can be used to efficiently produce approximate solutions to such a system for many values of the parameter at once. In this approach the parametrized system is first linearized, resulting in a large shifted linear system $(\boldsymbol{\mathsf{A}}-μ\boldsymbol{\mathsf{B}})\boldsymbol{\mathsf{y}}=\boldsymbol{\mathsf{d}}$. We formulate a left- and right-preconditioned rational Krylov GMRES method for shifted linear systems and show how the CORK framework can be used to speed up these methods in the setting of parametrized linear systems. Additionally, we show how to incorporate a right-hand side $\mathsf{b}(μ)$ that also depends on the parameter, how to choose the shifts to steer convergence and how to allow for inexact solves at these shifts throughout the iterations. As an application we consider the 'frequency sweeping' of Helmholtz scattering problems through the Boundary Element Method (BEM), enabled via an efficient representation of the dense but data-sparse wavenumber-dependent system matrix.

Problem

• parametrized linear systems
• frequency sweeping
• Helmholtz scattering
• Boundary Element Method
• nonlinear parameter dependence

Innovation

• Compact Rational Krylov (CORK)
• Parametrized Linear Systems
• Shifted Linear Systems
• Frequency Sweeping
• Boundary Element Method (BEM)