動的再構成可能ロボットシステムに関する研究 : 第9報, 自己組織化分散知識システムの定量解析および最適形態

The cellular robotic system (CEBOT), which has been studied by the authors, can carry out a variety of tasks by an optimal structurization. CEBOT consists of various kinds of components called "cells". A cell, which has a basic function and its own knowledge base and database, forms a group called a "cellular robotic society". In order to organize many kinds of structures for task execution and the cellular robotic society, each cell should genetically have algorithms similar to a living creature. This paper deals with a self-organizing method based on the genetic algorithm proposed by the authors. We report system organization methods (horizontal-and vertical-organization), and discuss how to organize the system effectively by using the proposed numerical analysis. As one example of the self-organizing system, a self-organizing distributed intelligence system based on the concept of CEBOT is reported. Some simulations of the self-organization and numerical analyzing results are also reported.