skip to main content
Tipo de recurso Mostra resultados com: Mostra resultados com: Índice

Global planning of whole-body manipulation by humanoid robot based on transition graph of object motion and contact switching

Murooka, Masaki ; Ueda, Ryohei ; Nozawa, Shunichi ; Kakiuchi, Yohei ; Okada, Kei ; Inaba, Masayuki

Advanced robotics, 2017-03, Vol.31 (6), p.322-340 [Periódico revisado por pares]

Taylor & Francis

Texto completo disponível

Citações Citado por
  • Título:
    Global planning of whole-body manipulation by humanoid robot based on transition graph of object motion and contact switching
  • Autor: Murooka, Masaki ; Ueda, Ryohei ; Nozawa, Shunichi ; Kakiuchi, Yohei ; Okada, Kei ; Inaba, Masayuki
  • Assuntos: Automation ; Contact ; graph search ; Graphical representations ; Graphs ; Humanoid ; humanoid robot ; manipulation planning ; quadratic programming ; Robots ; Searching ; Strategy ; Whole-body manipulation
  • É parte de: Advanced robotics, 2017-03, Vol.31 (6), p.322-340
  • Notas: ObjectType-Article-1
    SourceType-Scholarly Journals-1
    ObjectType-Feature-2
    content type line 23
  • Descrição: Humanoid robot has the potential to achieve various manipulation strategies such as lifting, pushing, pivoting and rolling. Deciding which strategies to apply for the object manipulation is difficult because the appropriate manipulation motion varies greatly depending on the object property. We propose the planning method of whole-body manipulation for various manipulati on strategies. In order to plan the manipulation in the global scope, we generate the graph representing all the possible object poses and operation contacts. By searching the feasible path in the graph, we plan the sequence of object states satisfying the kinematics and dynamics conditions. Finally, the robot postures for the whole-body manipulation are generated from the object states. The proposed method is efficient because the tractable object configuration is first planned in the global scope and then the complicated robot configuration is calculated. We show the effectiveness of the proposed method by planning the humanoid robot motion for various manipulation strategies automatically from the object and robot models.
  • Editor: Taylor & Francis
  • Idioma: Inglês

Buscando em bases de dados remotas. Favor aguardar.