Development of Internal Models for Communication of A Partner Robot based on Computational Intelligence

This paper discusses development of internal models of a partner robot for communication with a human based on computational intelligence. The mobile robot used as a partner robot must communicate with a human and make decisions through the facing environment. Therefore, the robot must perceive and understand the environment by the robot itself. In this paper, we propose a learning method using the multiple value functions of place and emotion. We also propose the method of interaction based on a fuzzy spiking neural network and a steady-state genetic algorithm. Furthermore, we show experimental results of the partner robot based on proposal method.