A Unified Approach to Controllers Design for Uncertain Systems

摘要

A model of uncertain systems based on the concept of probabilistic metrics is presented. It may be considered as a generalization of probabilistic, fuzzy, set membership point to set or inequality descriptions of uncertainty. Instead of a basic model of input-output, form, an extended model which maps elements of an input space to elements of a proper Menger space is proposed. A method of the feedback control design is an extension of Drenick's, point to set approach and uses fixed point theorem proved for special contraction mappings. The conditions for a controller which assures the reachability of satisfactory neighborhood of the desired output are given.