Generalized Sylvester Matrix Equations In Control Systems Theory
Duan Guang-Ren
2005-01-01
Abstract:A type of high-order generalized Sylvester matrix equations are concerned. This type of equations include the first- and second-order generalized Sylvester matrix equations, the normal Sylvester matrix equations of arbitrary orders and the well-known continuous- and discrete-time Lyapunov matrix equations all as special cases. It is first demonstrated that many problems in linear control systems theory are closely related with this type of generalized Sylvester matrix equations. Then, a condition for solvability of this type of generalized Sylvester matrix equations is given in terms of the newly defined concept of F-coprimeness of polynomial matrices, and for both cases of F being arbitrary and being in Jordan form, general complete parametric solutions to this type of equations are established, which give all the degree of freedom and may also allow the matrices to be chosen as a part of the degree of design freedom. Finally, by specifying the general parametric solutions, explicit and analytical solutions to the series of normal Sylvester matrix equations, including the continuous- and discrete-time Lyapunov matrix equations, are also obtained. It is pointed out, and partially demonstrated as well, that based on such general complete parametric solutions to these types of generalized Sylvester matrix. equations, general Sylvester parametric approaches for control system designs can be proposed, which are effective, simple and convenient, and also very powerful in dealing with multi-objective design requirements.