User Supplied Information in the Design of Linear Quadratic Gaussian Self-Tuning Controllers

Abstract

Abstract Practical implementation of LQG-selftuners can be substantially facilitated by CAD support. Moreover, user supplied information, though it is usually uncertain and sometimes seemingly irrelevant, can improve controller performance and start. The paper deals with translation of the user's knowledge expressed in terms of his every-day language into a well grounded mathematical problem formulation. Available information supplied by the user and corresponding mathematical tools of its exploitation are summarized and overall look at the state of art given.