A Riemannian Optimization Approach for Computing Low-Rank Solutions of Lyapunov Equations
Authors:
- Bart VandereyckenBVBart Vandereycken
- Stefan VandewalleSVStefan Vandewalle
- 1 January 2010
- journal article
- research article
- Published by Society for Industrial & Applied Mathematics (SIAM) in SIAM Journal on Matrix Analysis and Applications
- Vol. 31 (5) , 2553-2579
Abstract
We propose a new framework based on optimization on manifolds to approximate the solution of a Lyapunov matrix equation by a low-rank matrix. The method minimizes the error on the Riemannian manifold of symmetric positive semidefinite matrices of fixed rank. We detail how objects from differential geometry, like the Riemannian gradient and Hessian, can be efficiently computed for this manifold. As a minimization algorithm we use the Riemannian trust-region method of [P.-A. Absil, C. Baker, and K. Gallivan, Found. Comput. Math., 7 (2007), pp. 303–330] based on a second-order model of the objective function on the manifold. Together with an efficient preconditioner, this method can find low-rank solutions with very little memory. We illustrate our results with numerical examples.Keywords
This publication has 38 references indexed in Scilit:
- HSL_MI20: An efficient AMG preconditioner for finite element problems in 3DInternational Journal for Numerical Methods in Engineering, 2009
- A Geometric Newton Method for Oja's Vector FieldNeural Computation, 2009
- Newton methods for nonsmooth convex minimization: connections among -Lagrangian, Riemannian Newton and SQP methodsMathematical Programming, 2005
- Numerical solution of saddle point problemsActa Numerica, 2005
- Efficient numerical solution of the LQR‐problem for the heat equationProceedings in Applied Mathematics and Mechanics, 2004
- Existence of a low rank or ℋ︁‐matrix approximant to the solution of a Sylvester equationNumerical Linear Algebra with Applications, 2004
- A nonlinear programming algorithm for solving semidefinite programs via low-rank factorizationMathematical Programming, 2003
- A variance propagation algorithm for stochastic heat and mass transfer problems in food processesInternational Journal for Numerical Methods in Engineering, 2001
- Principal component analysis in linear systems: Controllability, observability, and model reductionIEEE Transactions on Automatic Control, 1981
- Algorithm 432 [C2]: Solution of the matrix equation AX + XB = C [F4]Communications of the ACM, 1972