Usage of linguistic variable concept for human operator modelling

Abstract

In many papers a human decision process is reproduced as a process of using minimum and product operators for research of a decision set on intersection goals and restrictions. Now mathematical psychology does not have any method for obtaining fuzzy goals and restrictions on an interval scale, thus using these operators for human decision modelling is limited. In the present paper a linguistic variable is assumed as some reflection of the real world in a person's consciousness. After the numbers have disappeared a human operator has to perform the operations with images (or words) of natural language only. The situation being thus, the concept of decision membership function may be introduced, which must be defined on a nominal scale of the combination of terms. The problem of an experiment planning for the definition of membership values having been considered, the analytical functions for their approximation are introduced. This results in decreasing of the dimension of the human operator modelling task in man-machine design. The results of our experiments indicate the possibility of model usage for evaluation of the operator's training level.