208104 可変神経振動子ネットワークによるロボットマニピュレータの運動軌道生成(OS17 ロボティクス・メカトロニクス2,オーガナイズド・セッション)

Abstract

In order to use robots effectively in the human life environment, an adaptive motion trajectory generation method based on sensor information is required. Central pattern generator to generate adaptive motion trajectory have been investigated by many researchers. However, to perform a number of different works and to return to works safely when large disturbances occurred have not been studied. So we propose an adaptive motion trajectory generation method by using structure changeable central pattern generator network. In this paper we report the experimental results of give disturbance to manipulator while it is stretching and bending arm, and swinging arm up and down.