Abstract
An iterative method for the low-rank approximate solution of a class of generalized Lyapunov equations is studied. At each iteration, a standard Lyapunov equation is solved using Galerkin projection with an extended Krylov subspace method. This Lyapunov equation is solved inexactly, thus producing a nonstationary iteration. Several theoretical and computational issues are discussed so as to make the iteration efficient. Numerical experiments indicate that this method is competitive vis-à-vis the current state-of-the-art methods, both in terms of computational times and storage needs.
Similar content being viewed by others
Notes
The value 0.3 is an estimate of the norm \(\Vert {\mathcal M}^{-1}{\mathcal N}\Vert _F\) for the first example heat1 in Sect. 5. In general, this norm is not near one for the problems considered here, where the rank of \(N_j \ll n\), and these matrices are extremely sparse.
We mention that similar slow results are obtained if one uses the tangential interpolation version of a rational Krylov method, as proposed in [7].
References
Benner, P., Breiten, T.: Low-rank methods for a class of generalized Lyapunov equations and related issues. Numerische Mathematik 124, 441–470 (2013)
Berman, A., Plemmons, R.J.: Nonnegative Matrices in the Mathematical Sciences. (3rd edn.) Academic Press, New York (1979). Reprinted by SIAM, Philadelphia (1994)
Damm, T.: Direct methods and ADI-preconditioned Krylov subspace methods for generalized Lyapunov equations. Numer. Linear Algebra Appl. 15, 853–871 (2008)
Damm, T., Hinrichsen, D.: Newton’s method for a rational matrix equation occurring in stochastic control. Linear Algebra Appl. 332–334, 81–109 (2001)
Dembo, R.S., Eisenstat, S.C., Steihaug, T.: Inexact Newton methods. SIAM J. Numer. Anal. 19, 400–408 (1982)
Druskin, V., Knizhnerman, L., Simoncini, V.: Analysis of the rational Krylov subspace and ADI methods for solving the Lyapunov equation. SIAM J. Numer. Anal. 49, 1875–1898 (2011)
Druskin, V., Simoncini, V., Zaslavsky, M.: Adaptive tangential interpolation in rational Krylov subspaces for MIMO model reduction data. SIAM J. Matrix Anal. Appl. 35, 476–498 (2014)
Fasi, M.: Weighted geometric mean of large-scale matrices: numerical analysis and algorithms. Master’s thesis. Department of Computer Science. Università di Bologna, Italy, March 2015
Frommer, A., Szyld, D.B.: On necessary conditions for convergence of stationary iterative methods for hermitian definite and semidefinite linear systems. Linear Algebra Appl. 453, 192–201 (2014)
Golub, G.H., Overton, M.L.: The convergence of inexact Chebyshev and Richardson iterative methods for solving linear systems. Numerische Mathematik 53, 571–593 (1988)
Hu, D.Y., Reichel, L.: Krylov subspace methods for the Sylvester equation. Linear Algebra Appl. 172, 283–313 (1992)
Knizhnerman, L., Simoncini, V.: Convergence analysis of the extended Krylov subspace method for the Lyapunov equation. Numerische Mathematik 118, 567–586 (2011)
Kressner, D., Tobler, C.: Low-rank tensor Krylov subspace methods for parametrized linear systems. SIAM J. Matrix Anal. Appl. 32, 1288–1316 (2011)
Lancaster, P., Tismenetsky, M.: The Theory of Matrices, 2nd edn. Academic Press, Orlando (1985)
Lanzkron, P.J., Rose, D.J., Szyld, D.B.: Convergence of nested classical iterative methods for linear systems. Numerische Mathematik 58, 685–702 (1991)
Matthies, H.G., Zander, E.: Solving stochastic systems with low-rank tensor compression. Linear Algebra Appl. 436, 3819–3838 (2012)
Nichols, N.K.: On the convergence of two-stage iterative processes for solving linear equations. SIAM J. Numer. Anal. 10, 460–469 (1973)
Penzl, T.: A cyclic low-rank Smith method for large sparse Lyapunov equations. SIAM J. Sci. Comput. 21, 1401–1418 (1999)
Shank, S.D.: Low-rank Solution Methods for Large-scale Linear Matrix Equations. PhD thesis, Department of Mathematics, Temple University, May 2014
Simoncini, V.. Computational methods for linear matrix equations. Preprint, Università di Bologna, March 2013, Revised January 2014. To appear in SIAM Review
Simoncini, V.: A new iterative method for solving large-scale Lyapunov matrix equations. SIAM J. Sci. Comput. 29, 1268–1288 (2007)
Varga, R.S.: Matrix Iterative Analysis. Prentice-Hall, Englewood Cliffs (1962). Second Edition, revised and expanded. Springer, Berlin (2000)
Acknowledgments
We are grateful to Peter Benner and Tobias Breiten for providing us with all their MATLAB codes from [1]. We also thank the referees for their questions and comments which helped improve our presentation.
Author information
Authors and Affiliations
Corresponding author
Additional information
This version dated 7 October 2015. This research is supported in part by the U.S. National Science Foundation under Grants DMS-1115520 and DMS-1418882, and by the FARB12SIMO Grant of the Università di Bologna.
Rights and permissions
About this article
Cite this article
Shank, S.D., Simoncini, V. & Szyld, D.B. Efficient low-rank solution of generalized Lyapunov equations. Numer. Math. 134, 327–342 (2016). https://doi.org/10.1007/s00211-015-0777-7
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00211-015-0777-7