Методика построения многоуровневой модели оценивания сложного объекта

Abstract

Reducing the dimension of a model that reflects dozens of various properties of a complex object is solved by dividing them into groups of generalized properties. Instead of an experienced or expert approach to the formation of general-ized properties, a systematic approach to the formation of properties of the upper level of generality is applied. The essence of the approach lies in the application of the manifestation principle of object properties and its parts in inter-action with the external environment. The proposed method for extracting a set of common properties of a complex object is applicable not only to objects of different purposes, but also of different nature (artificial and natural), which makes it possible to apply it to a human-machine system. To link properties of different levels of generality, a logical-linguistic analysis of concepts corresponding to the indicators used is applied. Linguistic analysis consists in extracting essential features from the definitions of concepts, and logical analysis - in determining the generic relationship between the essential features of the compared indicators. The proposed method is demonstrated on the example of creating a structure of indicators of a tethered high-altitude platform based on a multicopter. The possibility of obtaining partial estimates of platforms in the absence of completeness of indicators characterizing them is noted.