Reduced Sensitivity Solutions to the Global Linearisation of the Constant Pole Assignment Problem

Abstract

The paper develops further the global linearisation framework of the pole assignment by constant output feedback, which has given a simple proof of the mp>n result (necessary and sufficient condition for arbitrary pole placement of a generic p-input, m-output and n-state system) and has introduced a computational framework for finding output feedbacks. Although the existing solution is in closed form it has the disadvantage that in order to approximate the desired pole polynomial we have to approach the degenerate gain (singularity of the feedback configuration) used for the linearisation. Here the original method is improved by using higher order terms in the original method which guarantees some distance from the degenerate point.