Modelo Não Linear Neural de um Robô de Cinco Graus de Liberdade (Nonlinear Neural Model of a Robot Five Degrees of Freedom)

Jose Antonio Riul (riul@ct.ufpb.br)1, Ademar Gonçalves Costa Junior (ademar.costa@ifpb.edu.br)2, Paulo Henrique Miranda Montenegro (paulo@ct.ufpb.br)1


1Universidade Federal da Paraíba
2Programa de Pós Graduação em Engenharia Mecânica

This paper appears in: Revista IEEE América Latina

Publication Date: April 2016
Volume: 14,   Issue: 4 
ISSN: 1548-0992


Abstract:
The aim of this work is to evaluate the effectiveness of a nonlinear neural modeling of two links of an electromechanical manipulator robot five degrees of freedom (5 DOF). The robot consists of five rotational joints, four links and a claw. Five DC motors are used for driving the robot. The motion transfer of the actuators is accomplished through gear trains. The measurements of angular positions of the joints are made by potentiometers. Robot models are obtained using Newton - Euler or Lagrange equations, which are coupled and nonlinear. In this paper, a nonlinear neural model is obtained, using the excitations and responses data collected from the links 1 and 2 of the robot. Finally, results are presented using the model obtained.

Index Terms:
Robotic, Nonlinear Identification, Artificial Neural Networks.   


Documents that cite this document
This function is not implemented yet.


[PDF Full-Text (340)]