Interpretation and modeling of emotions in the management of autonomous robots using a control paradigm based on a scheduling variable

摘要

The paper presents a technical introduction to psychological theories of emotions. It highlights a usable idea implemented in a number of recently developed computational systems of emotions, and the hypothesis that emotion can play the role of a scheduling variable in controlling autonomous robots. In the main part of this study, we outline our own computational system of emotion - xEmotion - designed as a key structural element in the developed target device, being an Intelligent System of Decision-making (ISD) for autonomous and robotic units. The ISD system has a cognitive architecture based on the principles of human psychology. The main purpose of building such a system is to prepare a framework for autonomous units used in system engineering (Kowalczuk and Czubenko, 2011; Czubenko et al., 2015). In particular, ISD is based on the concepts of cognitive psychology (in information processing) and motivation theory, which includes the system of needs (for decisionmaking). The xEmotion subsystem, however, focuses on modeling an alternative approach based on emotion. The xEmotion implementation covers aspects of somatic, appraisal and evolutionary theories of emotions using fuzzy sets. In this article, we also illustrate the core emotional behavior of the ISD system using simulation. The first application is a user interface for identifying emotions and predicting human behavior. The second is an eSailor simulation, which illustrates the possible behavior of the xEmotion subsystem. The last is an xDriver simulation experiment, which is to prove the validity of the concept of using emotion-based systems, according to the SVC principle. In summary, we also discuss other possible applications of the xEmotion system.