Emotion model in robot assisted activity

Abstract

This paper discusses the robot assisted activity, which is physical and mental support for a human. The author proposes a function (mood congruent effect) that modifies the behavior of a robot based on its own emotion, and a control model (emotion-driven agent model) for the robot. The emotion-driven agent model has a recurrent type association memory and hierarchical knowledge. In the simulation, the agent showed a psychological defensive reaction.