Article in Proceedings INPROC-2005-20

BibliographyOubbati, Mohamed; Levi, Paul; Schanz, Michael: A Fixed-Weight RNN Dynamic Controller for Multiple Mobile Robots.
In: Proceedings of the 24th IASTED International Conference on MODELLING, IDENTIFICATION, AND CONTROL: MIC 2005; Innsbruck, Austria, February 16-18, 2005..
University of Stuttgart, Faculty of Computer Science, Electrical Engineering, and Information Technology.
pp. 277-282, english.
Austria: ACTA Press, February 2005.
Article in Proceedings (Conference Paper).
CorporationIASTED
CR-SchemaI.2.9 (Robotics)
KeywordsMobile Robot; Recurrent Neural Networks, Meta-Learning; Adaptive Control
Abstract

In this paper, we demonstrate the ability of a single fixed-weight RNN to act as a dynamic controller for several (here 3) distinct wheeled mobile robots, without exact knowledge about their dynamics parameters. The controller is properly trained to exhibit adaptive behaviour after its weights have been fixed. This capability is a natural consequence of prior meta-learning used recently in the area of RNNs.

ContactMohamed.Oubbati@informatik.uni-stuttgart.de
Department(s)University of Stuttgart, Institute of Parallel and Distributed Systems, Image Understanding
Project(s)COPS
SFB-627
Entry dateMarch 11, 2005
   Publ. Institute   Publ. Computer Science