クラスタリングに基づくスワームロボットシステムにおける群協調行動分析のための一手法

• 大倉 和博

  広島大学大学院工学研究院

• 保田 俊行

  広島大学大学院工学研究院

• 松村 嘉之

  信州大学繊維学部

• 森下 力

  広島大学大学院工学研究科

書誌事項

タイトル別名

• A Method of Analyzing Collective Behavior in a Swarm Robotic System Based on Clustering

抄録

Swarm robotic systems are a kind of multi-robot systems which consist of many homogeneous autonomous robots without any type of global controllers. Since a robotic swarm is controlled by an emergent way of many interactions with the other robots or an environment, such as a result of self-organization, robot learning or artificial evolution, no method has been known to grasp the collective behavior in a practical sense, according to the best of our knowledge. In this paper, based on this situation, a novel method for analyzing the collective behavior is proposed inspired by a technique in the field of complex networks. More specifically, a clustering technique based on the Modularity maximization is applied to a network associated with a robotic swarm which may be drawn by assuming that nodes are robots and links are informational connections with two nearest robots. We demonstrate that this technique enables us to find some robot subgroups each of which may have a certain functional role emerged in a robotic swarm as a community structure. The effectiveness of the proposed method is demonstrated by a series of computer simulations that deal with cooperative package pushing problems by ten autonomous robots.

収録刊行物

• 日本機械学会論文集Ｃ編

  日本機械学会論文集Ｃ編 78 (794), 3529-3540, 2012

  一般社団法人 日本機械学会

キーワード

• Complex Networks
• Clustering
• Collective Behavior
• Neural Network
• Swarm Robotics

詳細情報

• CRID: 1390001206386669184
• NII論文ID: 130002051590
• DOI: 10.1299/kikaic.78.3529
• ISSN: 18848354
• ISSN: 03875024
• Web Site: https://www.jstage.jst.go.jp/article/kikaic/78/794/78_3529/_pdf
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • Crossref
  • CiNii Articles
  • KAKEN
• 抄録ライセンスフラグ: 使用不可