セル集団の自己組織化によるロボットの形状モデリングと姿勢決定問題への適用

概要

論文の詳細を見る
This paper deals with a methodology of robot modeling for applying robot motion planning. Robot or robot component is modeled as a group of decision-making elements called cell. A cell is a kind of automaton which has the action objective to make uniform distribution as representing geometric shape of robot or robot component, and actions to achieve this objective. Process for making an arrangement of cell group by action of each cell is called self-organization. To give property of Self-Organization into cell, a potential function is defined on neighborhood of cell based on the concept of group theory. Each cell identically takes an action to go down its potential value. When minimal potential value is given to all cells in a group, self-organization is finished and a cell group is uniformly distributed as representing a robot component. As geometric shape of robot component is represented by a distribution of cell group, its position and direction correspond to robots configuration. This paper describes a modeling methodology of robot or robot component by cell group, and a way of Self-Organization is expounded. And results of simple case study executed to verify the effectiveness of selforganization are reported.
公益社団法人精密工学会の論文
1996-10-05