Learning algorithm improvements of a neural network based tuning method for robot controller

Author

Bossard, O. ; Kawamura, A.

Author_Institution

Dept. of Electr. & Comput. Eng., Yokohama Nat. Univ., Japan

Volume

2

fYear

1994

fDate

5-9 Sep 1994

Firstpage

1253

Abstract

A neural network based method has been proposed for control system´s gains tuning, but the network training is time consuming. Thus, several improvements of the classical backpropagation algorithm are investigated in order to speed-up the learning process. They are theoretically analyzed, and their effectiveness is confirmed with the application to the control of a nonlinear model of direct-drive two-axis robot manipulator. It is clarified that the convergence efficiency of the optimization process can be drastically increased by those improvements, resulting in a faster training of the network

Keywords

backpropagation; convergence; neural nets; robots; tuning; classical backpropagation algorithm; control system gains tuning; convergence efficiency; direct-drive two-axis robot manipulator; learning algorithm; neural network based tuning method; nonlinear model; robot controller; Backpropagation algorithms; Computer networks; Control systems; Digital filters; Electronic mail; Manipulators; Neural networks; Performance analysis; Robot control; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Electronics, Control and Instrumentation, 1994. IECON '94., 20th International Conference on

Conference_Location

Bologna

Print_ISBN

0-7803-1328-3

Type

conf

DOI

10.1109/IECON.1994.397973

Filename

397973