Neural-network-based control design: An LMI approach

Suttipan Limanond, Jennie Si

Research output: Contribution to journalArticle

111 Citations (Scopus)

Abstract

In this paper, we address a neural-network-based control design for a discrete-time nonlinear system. Our design approach is to approximate the nonlinear system with a multilayer perceptron of which the activation functions are of the sigmoid type symmetric to the origin. A linear difference inclusion representation is then established for this class of approximating neural networks and is used to design a state-feedback control law for the nonlinear system based on the certainty equivalence principle. The control design equations are shown to be a set of linear matrix inequalities where a convex optimization algorithm can be applied to determine the control signal. Further, the stability of the closed-loop is guaranteed in the sense that there exists a unique global attraction region in the neighborhood of the origin to which every trajectory of the closed-loop system converges. Finally, a simple example is presented so as to illustrate our control design procedure.

Original languageEnglish (US)
Pages (from-to)1422-1429
Number of pages8
JournalIEEE Transactions on Neural Networks
Volume9
Issue number6
DOIs
StatePublished - 1998

Fingerprint

LMI Approach
Control Design
Neural Networks
Neural networks
Nonlinear systems
Nonlinear Systems
Equivalence Principle
Nonlinear Discrete-time Systems
State Feedback Control
Activation Function
Signal Control
Perceptron
Convex Optimization
Closed-loop
Closed-loop System
Matrix Inequality
Multilayer
Linear Inequalities
Optimization Algorithm
Inclusion

Keywords

  • Linear matrix inequality
  • Multilayer perceptron
  • Nonlinear control
  • Stability analysis

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Theoretical Computer Science
  • Electrical and Electronic Engineering
  • Artificial Intelligence
  • Computational Theory and Mathematics
  • Hardware and Architecture

Cite this

Neural-network-based control design : An LMI approach. / Limanond, Suttipan; Si, Jennie.

In: IEEE Transactions on Neural Networks, Vol. 9, No. 6, 1998, p. 1422-1429.

Research output: Contribution to journalArticle

@article{4afd43c4e658456da686839f120c8906,
title = "Neural-network-based control design: An LMI approach",
abstract = "In this paper, we address a neural-network-based control design for a discrete-time nonlinear system. Our design approach is to approximate the nonlinear system with a multilayer perceptron of which the activation functions are of the sigmoid type symmetric to the origin. A linear difference inclusion representation is then established for this class of approximating neural networks and is used to design a state-feedback control law for the nonlinear system based on the certainty equivalence principle. The control design equations are shown to be a set of linear matrix inequalities where a convex optimization algorithm can be applied to determine the control signal. Further, the stability of the closed-loop is guaranteed in the sense that there exists a unique global attraction region in the neighborhood of the origin to which every trajectory of the closed-loop system converges. Finally, a simple example is presented so as to illustrate our control design procedure.",
keywords = "Linear matrix inequality, Multilayer perceptron, Nonlinear control, Stability analysis",
author = "Suttipan Limanond and Jennie Si",
year = "1998",
doi = "10.1109/72.728392",
language = "English (US)",
volume = "9",
pages = "1422--1429",
journal = "IEEE Transactions on Neural Networks and Learning Systems",
issn = "2162-237X",
publisher = "IEEE Computational Intelligence Society",
number = "6",

}

TY - JOUR

T1 - Neural-network-based control design

T2 - An LMI approach

AU - Limanond, Suttipan

AU - Si, Jennie

PY - 1998

Y1 - 1998

N2 - In this paper, we address a neural-network-based control design for a discrete-time nonlinear system. Our design approach is to approximate the nonlinear system with a multilayer perceptron of which the activation functions are of the sigmoid type symmetric to the origin. A linear difference inclusion representation is then established for this class of approximating neural networks and is used to design a state-feedback control law for the nonlinear system based on the certainty equivalence principle. The control design equations are shown to be a set of linear matrix inequalities where a convex optimization algorithm can be applied to determine the control signal. Further, the stability of the closed-loop is guaranteed in the sense that there exists a unique global attraction region in the neighborhood of the origin to which every trajectory of the closed-loop system converges. Finally, a simple example is presented so as to illustrate our control design procedure.

AB - In this paper, we address a neural-network-based control design for a discrete-time nonlinear system. Our design approach is to approximate the nonlinear system with a multilayer perceptron of which the activation functions are of the sigmoid type symmetric to the origin. A linear difference inclusion representation is then established for this class of approximating neural networks and is used to design a state-feedback control law for the nonlinear system based on the certainty equivalence principle. The control design equations are shown to be a set of linear matrix inequalities where a convex optimization algorithm can be applied to determine the control signal. Further, the stability of the closed-loop is guaranteed in the sense that there exists a unique global attraction region in the neighborhood of the origin to which every trajectory of the closed-loop system converges. Finally, a simple example is presented so as to illustrate our control design procedure.

KW - Linear matrix inequality

KW - Multilayer perceptron

KW - Nonlinear control

KW - Stability analysis

UR - http://www.scopus.com/inward/record.url?scp=0032207886&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032207886&partnerID=8YFLogxK

U2 - 10.1109/72.728392

DO - 10.1109/72.728392

M3 - Article

C2 - 18255820

AN - SCOPUS:0032207886

VL - 9

SP - 1422

EP - 1429

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

SN - 2162-237X

IS - 6

ER -