Numerical solutions to large-scale differential Lyapunov matrix equations

Abstract

Abstract. In the present paper, we consider large-scale differential Lyapunov matrix equations having a low rank constant term. We present two new approaches for the numerical resolution of such differential matrix equations. The first approach is based on the integral expression of the exact solution and an approximation method for the computation of the exponential of a matrix times a block of vectors. In the second approach, we first project the initial problem onto a block (or extended block) Krylov subspace and get a low-dimensional differential Lyapunov matrix equation. The latter differential matrix problem is then solved by the Backward Differentiation Formula method (BDF) and the obtained solution is used to build the low rank approximate solution of the original problem. The process being repeated until some prescribed accuracy is achieved. We give some new theoretical results and present some numerical experiments.

Global potential, topology, and pattern selection in a noisy stabilized Kuramoto–Sivashinsky equation

“…This is in a form known as the continuous Lyapunov equation, a type of linear equations which can be computed by a well-known algorithm (36)(37)(38). After solving Eq.…”

“…22, which is solved numerically by a standard routine in MatLab © , X = lyap(A, Q). This routine solves the continuous Lyapunov equation (37,38), AX + XA T + Q = 0. When formatting the codes, special care must be taken with the 1/L weight associated with the product of two matrices in Eq.…”

Stabilization of linear time‐varying reduced‐order models: A feedback controller approach

“…Balanced truncation requires solving matrix equations. Developing efficient approximate numerical algorithms for large matrix equations remains an active area of research 15‐19 . To avoid the high cost of balanced truncation, POD‐based 5,20‐25 and balanced POD (BPOD) 7,26 reduction methods are often used.…”

“…Developing efficient approximate numerical algorithms for large matrix equations remains an active area of research. [15][16][17][18][19] To avoid the high cost of balanced truncation, POD-based 5,[20][21][22][23][24][25] and balanced POD (BPOD) 7,26 reduction methods are often used. These approximate methods, however, in general lack the stability guarantee of balanced truncation, although a variety of heuristics methods are available to improve the stability properties of the ROMs.…”

A computational method for large‐scale differential symmetric Stein equation

“…Differential symmetric Stein equations (DSSE in short) have a fundamental role in linear control and filtering theory for discrete‐time large‐scale dynamical systems and other problems, see previous studies for more details. Recently, large‐scale differential matrix equations have drawn a great deal of attention, and some numerical approaches (projection methods) and splitting method have been proposed. The matrix Equation is also referred to as a discrete‐time Lyapunov equation.…”