Title :
A modified PD control of robot manipulator using neural network compensator
Author :
Huerta, Jos?© Antonio Heredia ; Yu, Wen
Author_Institution :
Seccion de Control Autom., CINVESTAV-IPN, Mexico City, Mexico
fDate :
6/21/1905 12:00:00 AM
Abstract :
In this paper a modified PD-like controller for robot manipulator is proposed. Nonlinear compensation terms are added to the PD controller. Since we assume the dynamic of the robot are unknown, RBF neural networks are used to estimate them. The neural compensator does not need off-line learning. The suggested learning laws are similar to the well-known backpropagation algorithm but with some additional terms. A Lyupunov-like analysis is used to derive these stable learning laws, as well as to assured the stability of the closed-loop system
Keywords :
closed loop systems; compensation; learning (artificial intelligence); manipulator dynamics; neurocontrollers; radial basis function networks; stability; two-term control; PD controller; closed-loop system; dynamics; learning laws; neural network; neurocontrol; nonlinear compensation; robot manipulator; stability; Asymptotic stability; Automatic control; Friction; Gravity; Manipulator dynamics; Neural networks; PD control; Robotics and automation; Robots; Transmission line matrix methods;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.832691
