Using time-discrete recurrent neural networks in nonlinear control

Author

Kolb, Thorsten ; Ilg, Winfried ; Wille, Jörg

Author_Institution

Brandenburg Univ. of Technol., Cottbus, Germany

Volume

2

fYear

1998

fDate

4-9 May 1998

Firstpage

1367

Abstract

We introduce a type of fully connected recurrent neural networks (RNN) with special mathematical features which allows one to determine its qualitative dynamical behaviour. Based on this family of RNN we describe a learning framework for the generation of trajectories with which we are able to solve adaptive control problems, which is illustrated by the realization of adaptive leg control of a six-legged walking machine

Keywords

adaptive control; learning (artificial intelligence); legged locomotion; motion control; neurocontrollers; nonlinear control systems; recurrent neural nets; adaptive control; learning; legged walking machine; mobile robot; neurocontrol; nonlinear control; time-discrete recurrent neural networks; Adaptive control; Backpropagation algorithms; Biological system modeling; Intelligent networks; Leg; Legged locomotion; Machine learning; Programmable control; Recurrent neural networks; Transfer functions;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on

Conference_Location

Anchorage, AK

ISSN

1098-7576

Print_ISBN

0-7803-4859-1

Type

conf

DOI

10.1109/IJCNN.1998.685974

Filename

685974