On the Synthesis of Insensitive Linear Feedback Control Systems

Publisher Website
Google Scholar
Abstract
In this paper analytic procedures are developed for the synthesis of linear feedback systems. The primary design objective is to render the system's transmission properties insensitive to plant variations. There is a fundamental conflict between sensitivity and the output error which is introduced by measurement noise. The synthesis procedures specify the compensation networks which achieve the best noise immunity under the conditions of insensitivity. For the class of systems considered, it has been held that feedback systems can realize independently specified transmission and sensitivity characteristics. It is found that this does not apply to systems which contain an incompletely known or time-varying plant. As a consequence, the mathematical formulation of the synthesis problem is based upon sets of variables which are descriptive of the capabilities of feedback systems. The design objectives are translated into the constrained minimization of a quadratic function of the variables. The analytic synthesis procedures are well suited for implementation by a digital computer.

This publication has 7 references indexed in Scilit: