A new application of the DIRECT model to the visual control of a robot

Author

Coronado, J.L. ; Guerrero-González, A.

Author_Institution

Univ. Politecnica de Cartagena, Spain

Volume

1

fYear

2000

fDate

2000

Firstpage

167

Abstract

In this paper a new application of the DIRECT neural algorithm on a robotic platform is presented. In (Guerrero et al., 1999) the first application of the DIRECT algorithm over a robotic platform was presented. The performance of this algorithm depends mainly on the neural maps which carry out the motor-spatial and spatial-motor transformations. In this paper, a new internal structure of the motor-spatial neural maps is proposed which allows reducing the required memory, enhancing the accuracy of the motor-spatial transformation, and reducing the learning times. The performance of this algorithm is analysed with several experiments on an experimental platform constituted by a robot arm and a stereohead

Keywords

learning (artificial intelligence); neurocontrollers; position control; robots; DIRECT neural algorithm; algorithm performance; experiments; learning times; motor-spatial neural maps; motor-spatial transformation; robot arm; robot visual control; spatial-motor transformation; stereohead; Algorithm design and analysis; Animals; Biological system modeling; Explosives; Humans; Neural networks; Performance analysis; Robot control; Robot kinematics; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man, and Cybernetics, 2000 IEEE International Conference on

Conference_Location

Nashville, TN

ISSN

1062-922X

Print_ISBN

0-7803-6583-6

Type

conf

DOI

10.1109/ICSMC.2000.884983

Filename

884983