行為に基づく環境モデリングのための移動ロボットの進化的行動獲得

• 山田 誠二

  東京工業大学大学院総合理工学研究科知能システム科学専攻

書誌事項

タイトル別名

• Evolutionary Learning of Behaviors for Action-Based Environment Modeling
• コウイ ニ モトヅク カンキョウ モデリング ノ タメ ノ イドウ ロボット ノ シンカテキ コウドウ カクトク

抄録

This paper describes an cvolutionary way to design behaviors of a mobile robot for recognizing environments. We have proposed AEM approach for a mobile robot to recognize environments. In AEM, a mobile robot acts in the environments and the action sequences are obtained. The action sequences are transformed into environment vectors which characterize the environments, and the robot identifies the environments with similarity between environment vectors. Suitable behaviors for AEM have been described by a human designer. However the design is very diflicult for them because of the huge search space. Thus we propose an evolutionary design method of such behaviors using genetic algorithm and make experiments in which a robot recognizes the environments with different structures. As results, we find out that our approach is promising to automatically acquire behaviors for AEM.

収録刊行物

• 人工知能

  人工知能 14 (5), 870-878, 1999-09-01

  一般社団法人 人工知能学会

キーワード

• action-based environment modeling
• a mobile robot
• behavior learning
• GA

詳細情報

• CRID: 1390004222625782144
• NII論文ID: 110002808760
• NII書誌ID: AN10067140
• ISSN: 09128085
• ISSN: 24358614
• ISSN: 21882266
• DOI: 10.11517/jjsai.14.5_870
• NDL書誌ID: 4841537
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I4841537
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • NDL
  • CiNii Articles
• 抄録ライセンスフラグ: 使用不可