Eigenvalue perturbation models for robust control

This paper presents a nonconservative description of the regions containing the eigenvalues of a diagonal matrix, perturbed by the set of all unknown, Euclidean norm bounded matrices. This is extended to real valued block diagonal matrices where each two-by-two block reveals a complex conjugate pair of eigenvalues. A weighting matrix allows one to specify the exact size of the perturbation to each of the eigenvalues. An identical result is obtained for real valued perturbations. This result is motivated by, and applied to, the modeling of frequency and damping perturbations in models of flexible structures. The resulting perturbation description fits within the established H/sub /spl infin// robust control framework and, in certain situations, is less conservative than the more standard additive or multiplicative perturbation models. >

[1]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[2]  G. Stewart Introduction to matrix computations , 1973 .

[3]  Roy S. Smith Eigenvalue perturbation models for flexible structures , 1991, [1991] Proceedings of the 30th IEEE Conference on Decision and Control.

[4]  Roy S. Smith,et al.  The design of H∞ controllers for an experimental non-collocated flexible structure problem , 1994, IEEE Trans. Control. Syst. Technol..

[5]  B. Morton,et al.  A Mu-Test for Robustness Analysis of a Real-Parameter Variation Problem , 1985, 1985 American Control Conference.

[6]  J.C. Doyle,et al.  Identification of flexible structures for robust control , 1990, IEEE Control Systems Magazine.