Identification of chaotic systems using a self-constructing recurrent neural network

Yen Ping Chen, Jeen-Shing Wang

Research output: Contribution to journalConference article

1 Citation (Scopus)

Abstract

This paper presents a self-constructing recurrent neural network (SCRNN) capable of building itself with a compact structure from input-output measurements for identification of chaotic systems. The proposed SCRNN is constituted by a static nonlinear network cascaded with a linear dynamic network. A unified learning algorithm consisting of two mechanisms, a hybrid weight initialization method and a parameter optimization method, has been developed for the structure and parameter identification. With this learning algorithm, the SCRNN is exempted from trial and error in structure initialization as well as parameterization. Computer simulations on discrete-time chaotic systems, including logistic and Henon mappings, validate that the proposed SCRNN is capable of capturing the dynamical behavior of chaotic systems with a compact network size.

Original languageEnglish
Pages (from-to)2150-2155
Number of pages6
JournalConference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
Volume3
Publication statusPublished - 2005 Dec 1
EventIEEE Systems, Man and Cybernetics Society, Proceedings - 2005 International Conference on Systems, Man and Cybernetics - Waikoloa, HI, United States
Duration: 2005 Oct 102005 Oct 12

Fingerprint

Recurrent neural networks
Chaotic systems
Learning algorithms
Nonlinear networks
Parameterization
Logistics
Identification (control systems)
Computer simulation

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

@article{e16ca4f3ae6d4a5eb6ac92a89106a39d,
title = "Identification of chaotic systems using a self-constructing recurrent neural network",
abstract = "This paper presents a self-constructing recurrent neural network (SCRNN) capable of building itself with a compact structure from input-output measurements for identification of chaotic systems. The proposed SCRNN is constituted by a static nonlinear network cascaded with a linear dynamic network. A unified learning algorithm consisting of two mechanisms, a hybrid weight initialization method and a parameter optimization method, has been developed for the structure and parameter identification. With this learning algorithm, the SCRNN is exempted from trial and error in structure initialization as well as parameterization. Computer simulations on discrete-time chaotic systems, including logistic and Henon mappings, validate that the proposed SCRNN is capable of capturing the dynamical behavior of chaotic systems with a compact network size.",
author = "Chen, {Yen Ping} and Jeen-Shing Wang",
year = "2005",
month = "12",
day = "1",
language = "English",
volume = "3",
pages = "2150--2155",
journal = "Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics",
issn = "1062-922X",

}

TY - JOUR

T1 - Identification of chaotic systems using a self-constructing recurrent neural network

AU - Chen, Yen Ping

AU - Wang, Jeen-Shing

PY - 2005/12/1

Y1 - 2005/12/1

N2 - This paper presents a self-constructing recurrent neural network (SCRNN) capable of building itself with a compact structure from input-output measurements for identification of chaotic systems. The proposed SCRNN is constituted by a static nonlinear network cascaded with a linear dynamic network. A unified learning algorithm consisting of two mechanisms, a hybrid weight initialization method and a parameter optimization method, has been developed for the structure and parameter identification. With this learning algorithm, the SCRNN is exempted from trial and error in structure initialization as well as parameterization. Computer simulations on discrete-time chaotic systems, including logistic and Henon mappings, validate that the proposed SCRNN is capable of capturing the dynamical behavior of chaotic systems with a compact network size.

AB - This paper presents a self-constructing recurrent neural network (SCRNN) capable of building itself with a compact structure from input-output measurements for identification of chaotic systems. The proposed SCRNN is constituted by a static nonlinear network cascaded with a linear dynamic network. A unified learning algorithm consisting of two mechanisms, a hybrid weight initialization method and a parameter optimization method, has been developed for the structure and parameter identification. With this learning algorithm, the SCRNN is exempted from trial and error in structure initialization as well as parameterization. Computer simulations on discrete-time chaotic systems, including logistic and Henon mappings, validate that the proposed SCRNN is capable of capturing the dynamical behavior of chaotic systems with a compact network size.

UR - http://www.scopus.com/inward/record.url?scp=27944436518&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=27944436518&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:27944436518

VL - 3

SP - 2150

EP - 2155

JO - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

JF - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

SN - 1062-922X

ER -