Recurrent fuzzy-neural network with fast learning algorithm for predictive control

Yancho Todorov, Margarita Terzyiska, Michail Petrov

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

5 Citations (Scopus)

Abstract

This paper presents a Takagi-Sugeno type recurrent fuzzy-neural network with a global feedback. To improve the predictions and to minimize the possible model oscillations, a hybrid learning procedure based on Gradient descent and the fast converging Gauss-Newton algorithms, is designed. The model performance is evaluated in prediction of two chaotic time series - Mackey-Glass and Rossler. The proposed recurrent fuzzy-neural network is coupled with analytical optimization approach in a Model Predictive Control scheme. The potentials of the obtained predictive controller are demonstrated by simulation experiments to control a nonlinear Continuous Stirred Tank Reactor.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning, ICANN 2013 -
Subtitle of host publication23rd International Conference on Artificial Neural Networks
PublisherSpringer
Pages459-466
ISBN (Electronic)978-3-642-40728-4
ISBN (Print)978-3-642-40727-7
DOIs
Publication statusPublished - 8 Oct 2013
MoE publication typeA4 Article in a conference publication
Event23rd International Conference on Artificial Neural Networks, ICANN 2013 - Sofia, Bulgaria
Duration: 10 Sep 201313 Sep 2013

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume8131
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference23rd International Conference on Artificial Neural Networks, ICANN 2013
CountryBulgaria
CitySofia
Period10/09/1313/09/13

Fingerprint

Predictive Control
Recurrent neural networks
Fuzzy neural networks
Fuzzy Neural Network
Learning algorithms
Fast Algorithm
Learning Algorithm
Hybrid Learning
Gauss-Newton
Chaotic Time Series
Prediction
Gradient Descent
Model predictive control
Model Predictive Control
Performance Model
Reactor
Simulation Experiment
Time series
Oscillation
Feedback

Keywords

  • Gauss-Newton method
  • Gradient descent
  • momentum learning
  • optimization
  • predictive control
  • recurrent fuzzy-neural networks
  • Takagi-Sugeno

Cite this

Todorov, Y., Terzyiska, M., & Petrov, M. (2013). Recurrent fuzzy-neural network with fast learning algorithm for predictive control. In Artificial Neural Networks and Machine Learning, ICANN 2013 - : 23rd International Conference on Artificial Neural Networks (pp. 459-466). Springer. Lecture Notes in Computer Science, Vol.. 8131 https://doi.org/10.1007/978-3-642-40728-4_58
Todorov, Yancho ; Terzyiska, Margarita ; Petrov, Michail. / Recurrent fuzzy-neural network with fast learning algorithm for predictive control. Artificial Neural Networks and Machine Learning, ICANN 2013 - : 23rd International Conference on Artificial Neural Networks. Springer, 2013. pp. 459-466 (Lecture Notes in Computer Science, Vol. 8131).
@inproceedings{8d94c954c40e40c394ed2199149f1c8d,
title = "Recurrent fuzzy-neural network with fast learning algorithm for predictive control",
abstract = "This paper presents a Takagi-Sugeno type recurrent fuzzy-neural network with a global feedback. To improve the predictions and to minimize the possible model oscillations, a hybrid learning procedure based on Gradient descent and the fast converging Gauss-Newton algorithms, is designed. The model performance is evaluated in prediction of two chaotic time series - Mackey-Glass and Rossler. The proposed recurrent fuzzy-neural network is coupled with analytical optimization approach in a Model Predictive Control scheme. The potentials of the obtained predictive controller are demonstrated by simulation experiments to control a nonlinear Continuous Stirred Tank Reactor.",
keywords = "Gauss-Newton method, Gradient descent, momentum learning, optimization, predictive control, recurrent fuzzy-neural networks, Takagi-Sugeno",
author = "Yancho Todorov and Margarita Terzyiska and Michail Petrov",
year = "2013",
month = "10",
day = "8",
doi = "10.1007/978-3-642-40728-4_58",
language = "English",
isbn = "978-3-642-40727-7",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "459--466",
booktitle = "Artificial Neural Networks and Machine Learning, ICANN 2013 -",
address = "Germany",

}

Todorov, Y, Terzyiska, M & Petrov, M 2013, Recurrent fuzzy-neural network with fast learning algorithm for predictive control. in Artificial Neural Networks and Machine Learning, ICANN 2013 - : 23rd International Conference on Artificial Neural Networks. Springer, Lecture Notes in Computer Science, vol. 8131, pp. 459-466, 23rd International Conference on Artificial Neural Networks, ICANN 2013, Sofia, Bulgaria, 10/09/13. https://doi.org/10.1007/978-3-642-40728-4_58

Recurrent fuzzy-neural network with fast learning algorithm for predictive control. / Todorov, Yancho; Terzyiska, Margarita; Petrov, Michail.

Artificial Neural Networks and Machine Learning, ICANN 2013 - : 23rd International Conference on Artificial Neural Networks. Springer, 2013. p. 459-466 (Lecture Notes in Computer Science, Vol. 8131).

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

TY - GEN

T1 - Recurrent fuzzy-neural network with fast learning algorithm for predictive control

AU - Todorov, Yancho

AU - Terzyiska, Margarita

AU - Petrov, Michail

PY - 2013/10/8

Y1 - 2013/10/8

N2 - This paper presents a Takagi-Sugeno type recurrent fuzzy-neural network with a global feedback. To improve the predictions and to minimize the possible model oscillations, a hybrid learning procedure based on Gradient descent and the fast converging Gauss-Newton algorithms, is designed. The model performance is evaluated in prediction of two chaotic time series - Mackey-Glass and Rossler. The proposed recurrent fuzzy-neural network is coupled with analytical optimization approach in a Model Predictive Control scheme. The potentials of the obtained predictive controller are demonstrated by simulation experiments to control a nonlinear Continuous Stirred Tank Reactor.

AB - This paper presents a Takagi-Sugeno type recurrent fuzzy-neural network with a global feedback. To improve the predictions and to minimize the possible model oscillations, a hybrid learning procedure based on Gradient descent and the fast converging Gauss-Newton algorithms, is designed. The model performance is evaluated in prediction of two chaotic time series - Mackey-Glass and Rossler. The proposed recurrent fuzzy-neural network is coupled with analytical optimization approach in a Model Predictive Control scheme. The potentials of the obtained predictive controller are demonstrated by simulation experiments to control a nonlinear Continuous Stirred Tank Reactor.

KW - Gauss-Newton method

KW - Gradient descent

KW - momentum learning

KW - optimization

KW - predictive control

KW - recurrent fuzzy-neural networks

KW - Takagi-Sugeno

UR - http://www.scopus.com/inward/record.url?scp=84884929578&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-40728-4_58

DO - 10.1007/978-3-642-40728-4_58

M3 - Conference article in proceedings

SN - 978-3-642-40727-7

T3 - Lecture Notes in Computer Science

SP - 459

EP - 466

BT - Artificial Neural Networks and Machine Learning, ICANN 2013 -

PB - Springer

ER -

Todorov Y, Terzyiska M, Petrov M. Recurrent fuzzy-neural network with fast learning algorithm for predictive control. In Artificial Neural Networks and Machine Learning, ICANN 2013 - : 23rd International Conference on Artificial Neural Networks. Springer. 2013. p. 459-466. (Lecture Notes in Computer Science, Vol. 8131). https://doi.org/10.1007/978-3-642-40728-4_58