Nonlinear tracking control via dynamic neural networks

被引：0

作者：

Song, YD ^{[1
]}

Liao, XH ^{[1
]}

Castelli, V ^{[1
]}

机构：

[1] N Carolina Agr & Tech State Univ, Dept Elect Engn, Greensboro, NC 27411 USA

来源：

CCCT 2003, VOL6, PROCEEDINGS: COMPUTER, COMMUNICATION AND CONTROL TECHNOLOGIES: III | 2003年

关键词：

system stability; tracking control; neural network; nonlinear system;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper presents an approach to tracking control of a class of nonlinear systems via dynamic neural networks. The control scheme proposed here consists of two major components: nominal model based control and recurrent network based compensation: It is shown that off-line training of the network is not needed and the stability is always ensured. The novelty of the proposed control scheme also lies its simplicity and effectiveness in dealing with nonlinearities and uncertainties in the system under consideration.

引用

页码：327 / 332

页数：6

共 7 条

[1] LEARNING LONG-TERM DEPENDENCIES WITH GRADIENT DESCENT IS DIFFICULT [J].

BENGIO, Y ;

SIMARD, P ;

FRASCONI, P .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1994, 5 (02) :157-166

[2] Learning long-term dependencies in NARX recurrent neural networks [J].

Lin, TN ;

Horne, BG ;

Tino, P ;

Giles, CL .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1996, 7 (06) :1329-1338

[3]

ROVITHAKIS GA, 1995, IEEE T SYST MAN CYB, V25, P1578, DOI 10.1109/21.478444

[4]

SEIDL D, 1991, P INT JOINT C NEUR N, V2, P709

[5] ON THE COMPUTATIONAL POWER OF NEURAL NETS [J].

SIEGELMANN, HT ;

SONTAG, ED .

JOURNAL OF COMPUTER AND SYSTEM SCIENCES, 1995, 50 (01) :132-150

[6]

[No title captured]

[7]

[No title captured]

← 1 →