Robust neural network proportional tracking controller with guaranteed global stability

Q. Song, M.J. Grimble

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

A robust neural network is proposed for use with a proportional fixed control scheme for robot control systems. A stability analysis is included based on sector theory. A special normalized learning algorithm is used to train the neural network, which eliminates the need for a bounded regression signal being input to the system. Furthermore, an adaptive dead zone scheme is employed to enhance the robustness of the control system against disturbances. A complete stability and convergence proof is included. The selection of the dead zone does not require knowledge of the upper bound of the disturbance, which is usually unknown for the robot control system. Simulation results are presented to demonstrate the effectiveness of the proposed robust control algorithm.

Conference

ConferenceIEEE International Symposium on Intelligent Control
CountryUnited States
CityHouston
Period5/10/038/10/03

Fingerprint

Neural networks
Control systems
Controllers
Robots
Robust control
Robustness (control systems)
Learning algorithms

Keywords

  • neural network
  • adaptive dead zone
  • conic sector
  • robot control
  • tracking controller
  • global stability

Cite this

Song, Q., & Grimble, M. J. (2003). Robust neural network proportional tracking controller with guaranteed global stability. 34-39. Paper presented at IEEE International Symposium on Intelligent Control , Houston, United States. https://doi.org/10.1109/ISIC.2003.1253910
Song, Q. ; Grimble, M.J. / Robust neural network proportional tracking controller with guaranteed global stability. Paper presented at IEEE International Symposium on Intelligent Control , Houston, United States.6 p.
@conference{8c38337d6aa740019b48132f4da25e42,
title = "Robust neural network proportional tracking controller with guaranteed global stability",
abstract = "A robust neural network is proposed for use with a proportional fixed control scheme for robot control systems. A stability analysis is included based on sector theory. A special normalized learning algorithm is used to train the neural network, which eliminates the need for a bounded regression signal being input to the system. Furthermore, an adaptive dead zone scheme is employed to enhance the robustness of the control system against disturbances. A complete stability and convergence proof is included. The selection of the dead zone does not require knowledge of the upper bound of the disturbance, which is usually unknown for the robot control system. Simulation results are presented to demonstrate the effectiveness of the proposed robust control algorithm.",
keywords = "neural network, adaptive dead zone , conic sector, robot control, tracking controller, global stability",
author = "Q. Song and M.J. Grimble",
year = "2003",
month = "10",
doi = "10.1109/ISIC.2003.1253910",
language = "English",
pages = "34--39",
note = "IEEE International Symposium on Intelligent Control ; Conference date: 05-10-2003 Through 08-10-2003",

}

Song, Q & Grimble, MJ 2003, 'Robust neural network proportional tracking controller with guaranteed global stability' Paper presented at IEEE International Symposium on Intelligent Control , Houston, United States, 5/10/03 - 8/10/03, pp. 34-39. https://doi.org/10.1109/ISIC.2003.1253910

Robust neural network proportional tracking controller with guaranteed global stability. / Song, Q.; Grimble, M.J.

2003. 34-39 Paper presented at IEEE International Symposium on Intelligent Control , Houston, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Robust neural network proportional tracking controller with guaranteed global stability

AU - Song, Q.

AU - Grimble, M.J.

PY - 2003/10

Y1 - 2003/10

N2 - A robust neural network is proposed for use with a proportional fixed control scheme for robot control systems. A stability analysis is included based on sector theory. A special normalized learning algorithm is used to train the neural network, which eliminates the need for a bounded regression signal being input to the system. Furthermore, an adaptive dead zone scheme is employed to enhance the robustness of the control system against disturbances. A complete stability and convergence proof is included. The selection of the dead zone does not require knowledge of the upper bound of the disturbance, which is usually unknown for the robot control system. Simulation results are presented to demonstrate the effectiveness of the proposed robust control algorithm.

AB - A robust neural network is proposed for use with a proportional fixed control scheme for robot control systems. A stability analysis is included based on sector theory. A special normalized learning algorithm is used to train the neural network, which eliminates the need for a bounded regression signal being input to the system. Furthermore, an adaptive dead zone scheme is employed to enhance the robustness of the control system against disturbances. A complete stability and convergence proof is included. The selection of the dead zone does not require knowledge of the upper bound of the disturbance, which is usually unknown for the robot control system. Simulation results are presented to demonstrate the effectiveness of the proposed robust control algorithm.

KW - neural network

KW - adaptive dead zone

KW - conic sector

KW - robot control

KW - tracking controller

KW - global stability

U2 - 10.1109/ISIC.2003.1253910

DO - 10.1109/ISIC.2003.1253910

M3 - Paper

SP - 34

EP - 39

ER -

Song Q, Grimble MJ. Robust neural network proportional tracking controller with guaranteed global stability. 2003. Paper presented at IEEE International Symposium on Intelligent Control , Houston, United States. https://doi.org/10.1109/ISIC.2003.1253910