Emotions simulation in methodology of autonomous adaptive control

Abstract

We offer a standpoint that emotions are a necessary mechanism for autonomous control systems. By an autonomous controlled object we understand an object, that is controlled by a control system which is its on-board subsystem. The control system performs learning and control in one process. We develop a methodology of autonomous adaptive control (AAC), that allows us to construct a control system for a given controlled object. As the control goals we take the controlled object survival and the knowledge accumulation. As a whole these goals bring maximization of the controlled object lifetime. We suggest an emotions modeling mechanism (EM). We give the description of its functions and their implementation in AAC methodology. These functions are: (1) a compulsion of the control system for activity; (2) an appreciation of the CO current state at its quality; (3) an appreciation of the formed patterns and their usefulness for control goals; (4) an influence on tempo and depth of reasoning of decision making in current state; (5) providing the decision making subsystem with emotional appraisals of the recognized patterns; (6) providing transfer of information on patterns of emotional appraisals simultaneously with information on the patterns while the organism interacts with another organism.