Fuzzy self-adaptive radial basis function neural network-based control of a seven-link redundant industrial manipulator

D. P. Nanayakkara Thrishantha, Keigo Watanabe, Kazuo Kiguchi, Kiyotaka Izumi

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

This paper proposes a method for the identification of dynamics and control of a multi-link industrial robot manipulator using Runge-Kutta-Gill neural networks (RKGNNs). RKGNNs are used to identify an ordinary differential equation of the dynamics of the robot manipulator. A structured function neural network (NN) with sub-networks to represent the components of the dynamics is used in the RKGNNs. The sub-networks consist of shape adaptive radial basis function (RBF) NNs. An evolutionary algorithm is used to optimize the shape parameters and the weights of the RBFNNs. Due to the fact that the RKGNNs can accurately grasp the changing rates of the states, this method can effectively be used for long-term prediction of the states of the robot manipulator dynamics. Unlike in conventional methods, the proposed method can even be used without input torque information because a torque network is part of the functional network. This method can be proposed as an effective option for the dynamics identification of manipulators with high degrees-of-freedom, as opposed to the derivation of dynamic equations and making additional hardware changes as in the case of statistical parameter identification such as linear least-squares method. Experiments were carried out using a seven-link industrial manipulator. The manipulator was controlled for a given trajectory, using adaptive fuzzy selection of nonlinear dynamic models identified previously. Promising experimental results are obtained to prove the ability of the proposed method in capturing nonlinear dynamics of a multi-link manipulator in an effective manner.

Original languageEnglish
Pages (from-to)17-43
Number of pages27
JournalAdvanced Robotics
Volume15
Issue number1
DOIs
Publication statusPublished - 2001
Externally publishedYes

Fingerprint

Industrial manipulators
Redundant manipulators
Manipulators
Neural networks
Torque
Robots
Industrial robots
Degrees of freedom (mechanics)
Ordinary differential equations
Evolutionary algorithms
Dynamic models
Identification (control systems)
Trajectories
Hardware

Keywords

  • Evolutionary optimization
  • Multi-link robot arms
  • Radial basis functions
  • Runge-Kutta-Gill neural networks

ASJC Scopus subject areas

  • Control and Systems Engineering

Cite this

Fuzzy self-adaptive radial basis function neural network-based control of a seven-link redundant industrial manipulator. / Nanayakkara Thrishantha, D. P.; Watanabe, Keigo; Kiguchi, Kazuo; Izumi, Kiyotaka.

In: Advanced Robotics, Vol. 15, No. 1, 2001, p. 17-43.

Research output: Contribution to journalArticle

Nanayakkara Thrishantha, D. P. ; Watanabe, Keigo ; Kiguchi, Kazuo ; Izumi, Kiyotaka. / Fuzzy self-adaptive radial basis function neural network-based control of a seven-link redundant industrial manipulator. In: Advanced Robotics. 2001 ; Vol. 15, No. 1. pp. 17-43.
@article{92d89c495328493b9a91e240052a0f72,
title = "Fuzzy self-adaptive radial basis function neural network-based control of a seven-link redundant industrial manipulator",
abstract = "This paper proposes a method for the identification of dynamics and control of a multi-link industrial robot manipulator using Runge-Kutta-Gill neural networks (RKGNNs). RKGNNs are used to identify an ordinary differential equation of the dynamics of the robot manipulator. A structured function neural network (NN) with sub-networks to represent the components of the dynamics is used in the RKGNNs. The sub-networks consist of shape adaptive radial basis function (RBF) NNs. An evolutionary algorithm is used to optimize the shape parameters and the weights of the RBFNNs. Due to the fact that the RKGNNs can accurately grasp the changing rates of the states, this method can effectively be used for long-term prediction of the states of the robot manipulator dynamics. Unlike in conventional methods, the proposed method can even be used without input torque information because a torque network is part of the functional network. This method can be proposed as an effective option for the dynamics identification of manipulators with high degrees-of-freedom, as opposed to the derivation of dynamic equations and making additional hardware changes as in the case of statistical parameter identification such as linear least-squares method. Experiments were carried out using a seven-link industrial manipulator. The manipulator was controlled for a given trajectory, using adaptive fuzzy selection of nonlinear dynamic models identified previously. Promising experimental results are obtained to prove the ability of the proposed method in capturing nonlinear dynamics of a multi-link manipulator in an effective manner.",
keywords = "Evolutionary optimization, Multi-link robot arms, Radial basis functions, Runge-Kutta-Gill neural networks",
author = "{Nanayakkara Thrishantha}, {D. P.} and Keigo Watanabe and Kazuo Kiguchi and Kiyotaka Izumi",
year = "2001",
doi = "10.1163/156855301750095569",
language = "English",
volume = "15",
pages = "17--43",
journal = "Advanced Robotics",
issn = "0169-1864",
publisher = "Taylor and Francis Ltd.",
number = "1",

}

TY - JOUR

T1 - Fuzzy self-adaptive radial basis function neural network-based control of a seven-link redundant industrial manipulator

AU - Nanayakkara Thrishantha, D. P.

AU - Watanabe, Keigo

AU - Kiguchi, Kazuo

AU - Izumi, Kiyotaka

PY - 2001

Y1 - 2001

N2 - This paper proposes a method for the identification of dynamics and control of a multi-link industrial robot manipulator using Runge-Kutta-Gill neural networks (RKGNNs). RKGNNs are used to identify an ordinary differential equation of the dynamics of the robot manipulator. A structured function neural network (NN) with sub-networks to represent the components of the dynamics is used in the RKGNNs. The sub-networks consist of shape adaptive radial basis function (RBF) NNs. An evolutionary algorithm is used to optimize the shape parameters and the weights of the RBFNNs. Due to the fact that the RKGNNs can accurately grasp the changing rates of the states, this method can effectively be used for long-term prediction of the states of the robot manipulator dynamics. Unlike in conventional methods, the proposed method can even be used without input torque information because a torque network is part of the functional network. This method can be proposed as an effective option for the dynamics identification of manipulators with high degrees-of-freedom, as opposed to the derivation of dynamic equations and making additional hardware changes as in the case of statistical parameter identification such as linear least-squares method. Experiments were carried out using a seven-link industrial manipulator. The manipulator was controlled for a given trajectory, using adaptive fuzzy selection of nonlinear dynamic models identified previously. Promising experimental results are obtained to prove the ability of the proposed method in capturing nonlinear dynamics of a multi-link manipulator in an effective manner.

AB - This paper proposes a method for the identification of dynamics and control of a multi-link industrial robot manipulator using Runge-Kutta-Gill neural networks (RKGNNs). RKGNNs are used to identify an ordinary differential equation of the dynamics of the robot manipulator. A structured function neural network (NN) with sub-networks to represent the components of the dynamics is used in the RKGNNs. The sub-networks consist of shape adaptive radial basis function (RBF) NNs. An evolutionary algorithm is used to optimize the shape parameters and the weights of the RBFNNs. Due to the fact that the RKGNNs can accurately grasp the changing rates of the states, this method can effectively be used for long-term prediction of the states of the robot manipulator dynamics. Unlike in conventional methods, the proposed method can even be used without input torque information because a torque network is part of the functional network. This method can be proposed as an effective option for the dynamics identification of manipulators with high degrees-of-freedom, as opposed to the derivation of dynamic equations and making additional hardware changes as in the case of statistical parameter identification such as linear least-squares method. Experiments were carried out using a seven-link industrial manipulator. The manipulator was controlled for a given trajectory, using adaptive fuzzy selection of nonlinear dynamic models identified previously. Promising experimental results are obtained to prove the ability of the proposed method in capturing nonlinear dynamics of a multi-link manipulator in an effective manner.

KW - Evolutionary optimization

KW - Multi-link robot arms

KW - Radial basis functions

KW - Runge-Kutta-Gill neural networks

UR - http://www.scopus.com/inward/record.url?scp=0035031136&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0035031136&partnerID=8YFLogxK

U2 - 10.1163/156855301750095569

DO - 10.1163/156855301750095569

M3 - Article

VL - 15

SP - 17

EP - 43

JO - Advanced Robotics

JF - Advanced Robotics

SN - 0169-1864

IS - 1

ER -