Analysis and Classification of Methods of Neural and Phase Logic for the Development of Parametric Identification Algorithms for Control Objects

Abstract

A general algorithm for identification of fibrous materials based on expert knowledge with an appropriate selection of the number and form of the membership functions of the fuzzy sets used in the model is determined. The basic methods of parametric identification of a control object is considered, the first of which is based on a structural model of the object while the second uses only a training sample.