Minimal indices cancellation and rank revealing factorizations for rational matrix functions

We study the problem of eliminating the minimal indices (the indices of a minimal polynomial basis of the null space) of a rational matrix function by multiplication with a suitably chosen invertible rational matrix function. We derive the class of all invertible factors that dislocate the minimal indices to certain zero locations and feature minimal McMillan degree. We impose additional conditions on the factor like being J-unitary, or J-inner, either with respect to the imaginary axis or to the unit circle, and characterize the classes of solutions. En route we extend the well-known rank revealing factorization of a constant matrix to rational matrix functions. The results are completely general and apply in particular to matrices which are polynomial, strictly proper or improper, rank deficient, with arbitrary poles and zeros including at infinity. All characterization are made by using descriptor realizations while the associated computations are performed by employing (constant) unitary transformations and standard reliable procedures for eigenvalue assignment or for solving Riccati equations.

[1]  P. Dooren,et al.  An improved algorithm for the computation of structural invariants of a system pencil and related geometric aspects , 1997 .

[3]  C. Oara,et al.  Generalized Riccati Theory and Robust Control. A Popov Function Approach , 1999 .

[4]  H. Rosenbrock,et al.  State-space and multivariable theory, , 1970 .

[5]  P. Dooren.  The generalized eigenstructure problem in linear system theory , 1981 .

[6]  Israel Gohberg,et al.  Partial pole and zero displacement by cascade connection , 1989 .

[7]  András Varga,et al.  Computation of general inner-outer and spectral factorizations , 2000, IEEE Trans. Autom. Control..

[8]  V. Belevitch,et al.  Classical network theory , 1968 .

[9]  P. Dooren.  Polynomial matrix factorizations via recursive pole-zero cancellation , 1990 .

[10]  T. Kailath,et al.  Properties of the system matrix of a generalized state-space system , 1978 .

[11]  Israel Gohberg,et al.  Factorizations of and extensions to J-unitary rational matrix functions on the unit circle , 1992 .

[12]  C. Oar.  Constructive solutions to spectral and inner-outer factorizations with respect to the disk , 2005 .

[13]  J. Vandewalle,et al.  On the factorization of a nonsingular rational matrix , 1975 .

[14]  H. Kimura.  ( J , J' )-lossless factorization based on conjugation , 1992 .

[15]  J. Vandewalle,et al.  On the irreducible cascade synthesis of a system with a real rational transfer matrix , 1977 .

[16]  Sabine Van Huffel,et al.  SLICOT—A Subroutine Library in Systems and Control Theory , 1999 .

[17]  Generalized eigenvalue problems , 1967 .

[18]  G. Groenewald.  INVERSION OF TOEPLITZ OPERATORS WITH RATIONAL SYMBOLS VIA FACTORIZATION AND REALIZATION , 1996 .

[19]  H. Kimura.  Chain-Scattering Approach to H∞Control , 1996 .

[20]  András Varga,et al.  Minimal Degree Coprime Factorization of Rational Matrices , 1999, SIAM J. Matrix Anal. Appl..

[21]  Jr. G. Forney,et al.  Minimal Bases of Rational Vector Spaces, with Applications to Multivariable Linear Systems , 1975 .

[22]  H. Kimura.  Conjugation, interpolation and model-matching in H ∞ , 1989 .

[23]  P. Dooren.  The Computation of Kronecker's Canonical Form of a Singular Pencil , 1979 .

[24]  L. Mirsky,et al.  The Theory of Matrices , 1961, The Mathematical Gazette.

[25]  Gene H. Golub,et al.  Matrix computations , 1983 .