A Decision Making System of Robots Introducing a Re-construction of Emotions Based on Their Own Experiences

Abstract

We already proposed a decision making system for autonomous robots, which is based on a Markovian emotional model proposed by Banik. In our proposed system, decision making of the robot is designed automatically by two processes that are the construction of emotions and optimization of behavior selection policy for emotions. In this study, we introduce an experiences-based reconstruction of emotions to our decision making system as an approach to an automatic construction of decision making online learning.