Model Identification of a Compound Technical Object in Conditions of High Prior Uncertainty

Abstract

Identification is a tool for modelling compound objects, which, due to complexity or insufficient knowledge, as well as many random factors cannot be investigated on the basis of traditional theoretical concepts. The types of uncertainty arising in the identification of complex technical objects are considered. The problems of using existing approaches are discussed. An approach to solving the identification problem based on fuzzy logic is proposed. The application of the proposed method when choosing a model of a complex object is illustrated. The solution to the problem of identifying complex technical objects by a set of criteria of various importance is considered. Examples are presented that explain the features of the proposed approach.