2A1-D14 人間型ロボットのための逐次学習による可動領域の獲得

• 原田 篤

  筑波大

• 鈴木 健嗣

  筑波大

書誌事項

タイトル別名

• 2A1-D14 Action-oriented learning of an operating space for a humanoid robot

抄録

This paper describes a new method of self-modeling based on action oriented learning to construct the operating space for a humanoid robot. The humanoid robot determines the operating space in the joint space by its own active behavior and iterative learning. We utilize the Gaussian Mixture Model (GMM) and Variational Bayesian (VB) learning to construct the operating space which is acquired by learning the boundary in the joint space. The robot obtains the operating space by iterative and on-line learning from the joint angles. We conduct several experiments in a real environment in order to construct the operating space.

収録刊行物

• ロボティクス・メカトロニクス講演会講演概要集

  ロボティクス・メカトロニクス講演会講演概要集 2009 (0), _2A1-D14_1-_2A1-D14_2, 2009

  一般社団法人 日本機械学会

キーワード

• Humanoid
• Self-modeling
• Variational Bayesian Learning

詳細情報

• CRID: 1390001205936278912
• NII論文ID: 110008698112
• DOI: 10.1299/jsmermd.2009._2a1-d14_1
• ISSN: 24243124
• Web Site: https://www.jstage.jst.go.jp/article/jsmermd/2009/0/2009__2A1-D14_1/_pdf
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • Crossref
  • CiNii Articles
• 抄録ライセンスフラグ: 使用不可