Sequential design of linear quadratic state regulators with prescribed eigenvalues and specified relative stability (Q2641264)

The authors consider time-invariant linear systems \[ \dot x(t)=Ax(t)+Bu(t),\quad x(0)=x_ 0,\quad x(t)\in R^ n,\quad u(t)\in R^ m \] (multi-input multi-output). The problem is to find a proper linear state-feedback -K which ensures that the closed-loop system A-BK has poles located in desired positions or within given regions. Such a feedback gain is not necessary unique, therefore a quadratic performance index \(J=\int^{\infty}_{0}[x(t)^ TQx(t)+u(t)^ TRu(t)]dt\) is to minimize, too. Known results and notions are collected in the paper. By means of the frequency-domain optimality conditions a sequential method is developed. The retention of poles assures that the poles lying in the regions of interest will not be assigned elsewhere again. Exact pole assignment and placement of poles in given regions are discussed. The method is demonstrated by an illustrative example.