Skip to Main content Skip to Navigation
Conference papers

Learning to grasp in unknown environment by reinforcement learning and shaping

Abstract : The purpose of this study is to propose a new tool to define the posture of a complete anthropomorphic arm model during grasping taking into account task and environment constraints. The developed model is based on a neural network architecture mixing both supervised and reinforcement learning. The task constraints are materialized by target points to be reached by the fingertips on the surface of the object to be grasped while environment constraints are represented by obstacles. With no prior information on the shape, position and number of obstacles, the model is able to find a suitable solution according to specified criteria. Simulation results are proposed and commented.
Document type :
Conference papers
Complete list of metadatas

https://archivesic.ccsd.cnrs.fr/sic_01798002
Contributor : Clara Galliano <>
Submitted on : Wednesday, May 23, 2018 - 9:51:16 AM
Last modification on : Wednesday, October 16, 2019 - 1:54:02 PM

Identifiers

Collections

Citation

Nasser Rezzoug, Philippe Gorce, Alexandre Abellard, Mohamed Moncef Ben Khelifa, Patrick Abellard. Learning to grasp in unknown environment by reinforcement learning and shaping. IEEE International Conference on Systems, Man and Cybernetics, IEEE, Oct 2006, Taipei, Taiwan. pp.4487-4492, ⟨10.1109/ICSMC.2006.384851⟩. ⟨sic_01798002⟩

Share

Metrics

Record views

140