MatrixEquations.jl

Solution of Lyapunov, Sylvester and Riccati matrix equations using Julia

Compatibility

Julia 1.8 and higher.

How to Install

pkg> add MatrixEquations
pkg> test MatrixEquations

About

This collection of Julia functions is an attemp to implement high performance numerical software to solve several classes of Lyapunov, Sylvester and Riccati matrix equations at a performance level comparable with efficient structure exploiting Fortran implementations, as those available in the Systems and Control Library SLICOT. This goal has been fully achieved for Lyapunov and Sylvester equation solvers, for which the codes for both real and complex data perform at practically same performance level as similar functions available in the MATLAB Control System Toolbox (which rely on SLICOT).

The available functions in the MatrixEquation.jl package cover both standard and generalized continuous and discrete Lyapunov, Sylvester and Riccati equations for both real and complex data. The functions for the solution of Lyapunov and Sylvester equations rely on efficient structure exploiting solvers for which the input data are in Schur or generalized Schur forms. A comprehensive set of Lyapunov and Sylvester operators has been implemented, which allow the estimation of condition numbers of these operators. The implementation of Riccati equation solvers employ orthogonal Schur vectors based methods and their extensions to linear matrix pencil based reduction approaches. The calls of all functions with adjoint (in complex case) or transposed (in real case) arguments are fully supported by appropriate computational algorithms, thus the matrix copying operations are mostly avoided.

The current version of the package includes the following functions:

Solution of Lyapunov equations

• lyapc Solution of the continuous Lyapunov equations AX+XA'+C = 0 and AXE'+EXA'+C = 0.
• lyapd Solution of discrete Lyapunov equations AXA'-X +C = 0 and AXA'-EXE'+C = 0.
• tlyapc Solution of the continuous T-Lyapunov equations A*X±X^⊤*A^⊤+C = 0.
• hlyapc Solution of the continuous H-Lyapunov equations A*X±X'*A'+C = 0.
• tulyapc! Solution of the continuous T-Lyapunov equations U*X^⊤+X*U^⊤ = C and U^⊤*X+X^⊤*U = C for U and X upper triangular.
• hulyapc! Solution of the continuous H-Lyapunov equations U*X'+X*U' = C and U'*X+X'*U = C for U and X upper triangular.
• plyapc Solution of the positive continuous Lyapunov equations AX+XA'+BB' = 0 and AXE'+EXA'+BB' = 0.
• plyapd Solution of the positive discrete Lyapunov equations AXA'-X +C = 0 and AXA'-EXE'+C = 0.

Solution of algebraic Riccati equations

• arec Solution of the continuous Riccati equations A'X+XA-XRX+Q = 0 and A'X+XA-(XB+S)R^(-1)(B'X+S')+Q = 0.
• garec Solution of the generalized continuous Riccati equation A'XE+E'XA-(A'XB+S)R^(-1)(B'XA+S')+Q = 0.
• ared Solution of the discrete Riccati equation A'XA - X - (A'XB+S)(R+B'XB)^(-1)(B'XA+S') + Q = 0.
• gared Solution of the generalized discrete Riccati equation A'XA - E'XE - (A'XB+S)(R+B'XB)^(-1)(B'XA+S') + Q = 0.

Solution of Sylvester equations and systems

• sylvc Solution of the (continuous) Sylvester equation AX+XB = C.
• sylvd Solution of the (discrete) Sylvester equation AXB+X = C.
• gsylv Solution of the generalized Sylvester equation AXB+CXD = E.
• sylvsys Solution of the Sylvester system of matrix equations AX+YB = C, DX+YE = F.
• dsylvsys Solution of the dual Sylvester system of matrix equations AX+DY = C, XB+YE = F.

Iterative solution of linear matrix equations

• lyapci Iterative solution of the continuous Lyapunov equations AX+XA'+C = 0 and AXE'+EXA'+C = 0.
• lyapdi Iterative solution of the discrete Lyapunov equations AXA'-X+C = 0 and AXA'-EXE'+C = 0.
• tlyapci Iterative solution of the continuous T-Lyapunov equations A*X+X^⊤*A^⊤ = C and A*X^⊤+X*A^⊤ = C.
• hlyapci Iterative solution of the continuous H-Lyapunov equations A*X+X'*A' = C and A*X'+X*A' = C.
• tulyapci Iterative solution of the continuous T-Lyapunov equations U*X^⊤+X*U^⊤ = C and U^⊤*X+X^⊤*U = C for U and X upper triangular.
• hulyapci Iterative solution of the continuous H-Lyapunov equations U*X'+X*U' = C and U'*X+X'*U = C for U and X upper triangular.
• sylvci Iterative solution of the (continuous) Sylvester equation AX+XB = C.
• sylvdi Iterative solution of the (discrete) Sylvester equation AXB+X = C.
• gsylvi Iterative solution of the generalized Sylvester equation AXB+CXD = E.
• gtsylvi Iterative solution of the generalized T-Sylvester equations ∑ A_i*X*B_i + ∑ C_j*X^⊤*D_j = E.
• ghsylvi Iterative solution of the generalized T-Sylvester equations ∑ A_i*X*B_i + ∑ C_j*X'*D_j = E.
• cgls The conjugate gradient method for nonsymmetric linear equations and least squares problems.

Norm, condition and separation estimation

• opnorm1 Computation of the 1-norm of a linear operator.
• opnorm1est Estimation of 1-norm of a linear operator.
• oprcondest Estimation of the reciprocal 1-norm condition number of a linear operator.
• opsepest Estimation of the 1-norm separation a linear operator.

The general solvers of Lyapunov and Sylvester equations rely on a set of specialized solvers for real or complex matrices in appropriate Schur forms. For testing purposes, a set of solvers for various linear matrix equations has been implemented, which employ the Kronecker-product expansion of the equations. These solvers are not recommended for large order matrices. Based on the conjugate gradient method to solve linear systems or least-squares problems, several iterative solvers have been implemented, which potentially are applicable to solve linear matrix equations with large order dense and sparse matrices. The norms, reciprocal condition numbers and separations can be estimated for a comprehensive set of predefined Lyapunov and Sylvester operators.
A complete list of implemented functions is available here.

Future plans

The collection of tools can be extended by adding new functionality, such as expert solvers, which additionally compute error bounds and condition estimates, or solvers for new classes of Riccati equations, as those arising in game-theoretic optimization problems. Further performance improvements are still possible (e.g., in some positive Lyapunov solvers by employing specially taylored solvers for the underlying particular Sylvester equations) or by employing blocking based variants of solvers for Lyapunov and Sylvester equations.