Emotional Model for Robotic System Using a Self-Organizing Map Combined with Markovian Model

Abstract

<div class=""abs_img""> <img src=""[disp_template_path]/JRM/abst-image/00270005/13.jpg"" width=""300"" /> Behavioral/emotional expression system</div> In our research, we have focused on investigating the application of brain-inspired technology by developing a robot with consciousness resembling that of a human being. The goal was to enhance intelligent behavior/emotion, and to facilitate communication between human beings and robots. We sought to increase the robot’s behavioral/emotional intelligence capabilities so that it could distinguish, adapt and react to changes in the environment. In this paper, we present a behavioral/emotional expression system designed to work automatically by two processes. The first is a classification of behavior and emotions by determining the winner node based on Self-Organizing Map (SOM) learning. For the second, we propose a stochastic emotion model based on Markov theory in which the probabilities of emotional state transition are updated with affective factors. Finally, we verified this model with a conscious behavior robot (Conbe-I), and confirmed the effectiveness of the proposed system with the experimental results in a realistic environment. </span>