Improved neural network control of inverted pendulums

Teréz A. Várkonyi, J. Tar, I. Rudas

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Nowadays, neural network controllers (NNCs) are getting more and more prevalent because they are able to handle unknown systems by learning them and adapt to their changing behaviour. The family of robust fixed point transformations (RFPT) has been partly developed to solve control tasks without knowing the exact parameters of a controlled system. When disturbances effect a plant or the neural network controller is not trained properly RFPT integrated to the controller is suitable to reduce the problems caused by the model approximation and make the controller robust to the unknown external forces. In this paper, a novel combination of neura networks and robust fixed point transformations is introduced to balance an inverted pendulum on the top of a cart of changing nominal position. The results show that the inaccuracies caused by the disturbances can be reduced significantly when RFPT is used in the control process.

Original languageEnglish
Pages (from-to)270-283
Number of pages14
JournalInternational Journal of Advanced Intelligence Paradigms
Volume5
Issue number4
DOIs
Publication statusPublished - 2013

Fingerprint

Neural Network Control
Inverted Pendulum
Pendulums
Neural networks
Controllers
Fixed point
Controller
Disturbance
Neural Networks
Unknown
Process Control
Categorical or nominal
Approximation

Keywords

  • Adaptive control
  • Inverted pendulum
  • Neural network
  • Neural network control
  • RFPT
  • Robust control
  • Robust fixed point transformations

ASJC Scopus subject areas

  • Computer Science(all)
  • Engineering(all)
  • Applied Mathematics

Cite this

Improved neural network control of inverted pendulums. / Várkonyi, Teréz A.; Tar, J.; Rudas, I.

In: International Journal of Advanced Intelligence Paradigms, Vol. 5, No. 4, 2013, p. 270-283.

Research output: Contribution to journalArticle

@article{03571d5f4d7b47b69a0b0bdce7fa8afc,
title = "Improved neural network control of inverted pendulums",
abstract = "Nowadays, neural network controllers (NNCs) are getting more and more prevalent because they are able to handle unknown systems by learning them and adapt to their changing behaviour. The family of robust fixed point transformations (RFPT) has been partly developed to solve control tasks without knowing the exact parameters of a controlled system. When disturbances effect a plant or the neural network controller is not trained properly RFPT integrated to the controller is suitable to reduce the problems caused by the model approximation and make the controller robust to the unknown external forces. In this paper, a novel combination of neura networks and robust fixed point transformations is introduced to balance an inverted pendulum on the top of a cart of changing nominal position. The results show that the inaccuracies caused by the disturbances can be reduced significantly when RFPT is used in the control process.",
keywords = "Adaptive control, Inverted pendulum, Neural network, Neural network control, RFPT, Robust control, Robust fixed point transformations",
author = "V{\'a}rkonyi, {Ter{\'e}z A.} and J. Tar and I. Rudas",
year = "2013",
doi = "10.1504/IJAIP.2013.058303",
language = "English",
volume = "5",
pages = "270--283",
journal = "International Journal of Advanced Intelligence Paradigms",
issn = "1755-0386",
publisher = "Inderscience Enterprises Ltd",
number = "4",

}

TY - JOUR

T1 - Improved neural network control of inverted pendulums

AU - Várkonyi, Teréz A.

AU - Tar, J.

AU - Rudas, I.

PY - 2013

Y1 - 2013

N2 - Nowadays, neural network controllers (NNCs) are getting more and more prevalent because they are able to handle unknown systems by learning them and adapt to their changing behaviour. The family of robust fixed point transformations (RFPT) has been partly developed to solve control tasks without knowing the exact parameters of a controlled system. When disturbances effect a plant or the neural network controller is not trained properly RFPT integrated to the controller is suitable to reduce the problems caused by the model approximation and make the controller robust to the unknown external forces. In this paper, a novel combination of neura networks and robust fixed point transformations is introduced to balance an inverted pendulum on the top of a cart of changing nominal position. The results show that the inaccuracies caused by the disturbances can be reduced significantly when RFPT is used in the control process.

AB - Nowadays, neural network controllers (NNCs) are getting more and more prevalent because they are able to handle unknown systems by learning them and adapt to their changing behaviour. The family of robust fixed point transformations (RFPT) has been partly developed to solve control tasks without knowing the exact parameters of a controlled system. When disturbances effect a plant or the neural network controller is not trained properly RFPT integrated to the controller is suitable to reduce the problems caused by the model approximation and make the controller robust to the unknown external forces. In this paper, a novel combination of neura networks and robust fixed point transformations is introduced to balance an inverted pendulum on the top of a cart of changing nominal position. The results show that the inaccuracies caused by the disturbances can be reduced significantly when RFPT is used in the control process.

KW - Adaptive control

KW - Inverted pendulum

KW - Neural network

KW - Neural network control

KW - RFPT

KW - Robust control

KW - Robust fixed point transformations

UR - http://www.scopus.com/inward/record.url?scp=84890822687&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890822687&partnerID=8YFLogxK

U2 - 10.1504/IJAIP.2013.058303

DO - 10.1504/IJAIP.2013.058303

M3 - Article

AN - SCOPUS:84890822687

VL - 5

SP - 270

EP - 283

JO - International Journal of Advanced Intelligence Paradigms

JF - International Journal of Advanced Intelligence Paradigms

SN - 1755-0386

IS - 4

ER -