Root clustering in parameter space (Q1188623)

By root clustering is meant that the eigenvalues of a matrix of the zeros of a polynomial lie in a prescribed region. The well known Routh Hurwitz criterion is a special case of a root clustering result, where the left half plane is the given region. Another special case is the theorem of Lyapunov, which states that the matrix A has all its eigenvalues in the left half plane if and only if the equation \(PA+A^*P=-Q\), has for any positive definite matrix Q a positive definite solution P. In this book general criteria for root clustering are presented. After the introduction and a review of classical results, like the Routh Hurwitz and the Lyapunov criterion, the author gives a description of his class of regions. This class is wide, and it contains most regions used in applications, like e.g. cones. In chapters 5 and 6 the root clustering criteria are presented. Finally in the last chapter the problem of designing a compensator such that the poles of the closed loop system are in a prescribed region is treated. The results in this section show great similarity with the recent results in \(H_{\infty}\) control design, and with the results obtained in the theory of spectral radii.