Recycling Krylov subspaces and reducing deflation subspaces for solving sequence of linear systems

Hussam Al Daas 1 Laura Grigori 1 Pascal Hénon 2 Philippe Ricoux 3
1 ALPINES - Algorithms and parallel tools for integrated numerical simulations
INSMI - Institut National des Sciences Mathématiques et de leurs Interactions, Inria de Paris, LJLL - Laboratoire Jacques-Louis Lions
Abstract : This paper presents deflation strategies related to recycling Krylov subspace methods for solving one or a sequence of linear systems of equations. Besides well-known strategies of deflation, Ritz- and harmonic Ritz-based deflation, we introduce an SVD-based deflation technique. We consider the recycling in two contexts, recycling the Krylov subspace between the cycles of restarts and recycling a deflation subspace when the matrix changes in a sequence of linear systems. Numerical experiments on real-life reservoir simulations demonstrate the impact of our proposed strategy.
Document type :
Reports
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download

https://hal.inria.fr/hal-01886546
Contributor : Hussam Al Daas <>
Submitted on : Tuesday, October 2, 2018 - 9:39:02 PM
Last modification on : Friday, September 20, 2019 - 4:34:04 PM
Long-term archiving on : Thursday, January 3, 2019 - 4:14:07 PM

File

RR-9206.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01886546, version 1

Citation

Hussam Al Daas, Laura Grigori, Pascal Hénon, Philippe Ricoux. Recycling Krylov subspaces and reducing deflation subspaces for solving sequence of linear systems. [Research Report] RR-9206, Inria Paris. 2018. ⟨hal-01886546⟩

Share

Metrics

Record views

181

Files downloads

193