Solving Geometrical Uncertainty of Mobile Manipulation by Natural Actor-Critic

Title
Solving Geometrical Uncertainty of Mobile Manipulation by Natural Actor-Critic
Authors
김병찬전창묵강성철류동석박신석
Keywords
Mobile manipulation,; opening a door; compliance control; reinforcement learning; pose error
Issue Date
2009-10
Publisher
URAI 2009
Abstract
This paper is a study on an adaptive method for opening a door using a mobile manipulator. Opening a door has been regarded as a geometry-oriented problem. This work treat the door opening as a force-oriented problem. Not assuming a door’s exact trajectory, only compliance control is executed with a simple command which directs a pulling direction. Then, the mobile manipulator adapts itself to the door trajectory, with bounded force condition. The main challenge of this research is to find an optimum compliance gain at each configuration of the manipulator, with position error of mobile base. To resolve this problem, we employed an adaptive method based on modern reinforcement learning. We simulated the learning process for a door opening task, then adopted the learned compliance gain policy to the opening a door execution using real mobile manipulator. The experimental results prove that the proposed adaptive strategy was successful.
URI
http://pubs.kist.re.kr/handle/201004/36555
Appears in Collections:
KIST Publication > Conference Paper
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE