セル集団の自己組織化を適用したロボットアームの連続的な姿勢変化の生成

Abstract

This paper deals with collision-free motion planning for an articulated robot arm. In this study, a link constructing an articulated robot arm is modeled as a group of element called cell. An articulated robot arm is represented by combining groups of the cell in each joint position. Each cell has the action objectives that represent a shape of link by distributing all cells as grid pattern in the collision-free space. Process of achieving this objective is called self-organization. Distribution of cell group obtained by self-organization is regarded as collision-free configuration of an articulated robot arm. In proposed approach, a problem of motion planning of an articulated robot arm is solved by integrating end-effector path planning from a given initial position until a target position and determination of configuration by using self-organization of cell group. This paper explains a method for planning end-effector paths, and then describes a method for determination of arm configuration for position of an end-effector. Finally, results of motion planning for a three-revolute-joint arm are reported to verify the effectiveness of proposed method.