Emotion Behavior Learning System for Autonomous Mobile Robot Based on Target Selection-Type Q-Learning

Abstract

In recent years, in the field of the entertainment robot and the nursing care robot, several methods for embedding human emotions to the robot are developed actively. We have considered a learning method of the emotion behaviors of an autonomous mobile robot by using Q-Learning method with the meta-parameters controlled by neuromodulators which are well known in the field of the brain science so far. In this paper, we propose a target selection-type Q-Learning method with the plural Q-values concerning the maximization and the minimization of rewards and punishments. We aim at the realization of a system that adaptively learns complicated emotion behaviors with the behavior selection based on the positive and negative evaluation according to the situation. Furthermore, we also report the result of an experiment by the computer simulation to confirm the efficiency of the proposed method.