General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization

I. Rudas, F. Pereszlenyi, J. Tar, J. F. Bito

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Recent development of adaptive and robust controllers for non-linear coupled systems is seriously inspired by new possibilities of hardware-realization. It concerns each kinds of artificial intelligence: the 'classical' knowledge-based systems (KBSs), the more modern artificial neural network- (ANN) and fuzzy sets- (FS) based formulations. While common and general mathematical roots of these approaches became transparent, in each concrete case complexity of the appropriate control rules, the number of the ANN processor units or fuzzy sets, complexity of the fuzzy rules of a satisfactory controller essentially depend on the generalized coordinates used for describing the system to be controlled. In this paper it is pointed out, that in a wide class of control tasks relatively simple kids of coordinate transformation can be applied to achieve effective simplification of the control rules and reduction of the number of the necessary processor units or fuzzy sets. As an example, dynamics of robot arms are considered, where the appropriate coordinate transformation is approximate diagonalization of the inertia matrix of the robot arms. It serves as a satisfactory basis for working out a general method and principles for dealing with the quadratic coupling not eliminated by the diagonalization. It shows strong robustness with respect to the unknown inertia of the manipulated body and easily can be realized by ANN or fuzzy rule-based solutions of very limited number of processor units or fuzzy rules. The method is demonstrated by results of numerical simulation for a concrete physical system. Effects of imprecise modeling of the mass of the work-piece carried by the robot are considered, too.

Original languageEnglish
Title of host publicationIECON Proceedings (Industrial Electronics Conference)
Editors Anon
PublisherPubl by IEEE
Pages67-72
Number of pages6
Volume1
ISBN (Print)0780308913
Publication statusPublished - 1993
EventProceedings of the 19th International Conference on Industrial Electronics, Control and Instrumentation - Maui, Hawaii, USA
Duration: Nov 15 1993Nov 18 1993

Other

OtherProceedings of the 19th International Conference on Industrial Electronics, Control and Instrumentation
CityMaui, Hawaii, USA
Period11/15/9311/18/93

Fingerprint

Fuzzy rules
Fuzzy sets
Robots
Neural networks
Controllers
Knowledge based systems
Artificial intelligence
Hardware
Computer simulation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Rudas, I., Pereszlenyi, F., Tar, J., & Bito, J. F. (1993). General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization. In Anon (Ed.), IECON Proceedings (Industrial Electronics Conference) (Vol. 1, pp. 67-72). Publ by IEEE.

General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization. / Rudas, I.; Pereszlenyi, F.; Tar, J.; Bito, J. F.

IECON Proceedings (Industrial Electronics Conference). ed. / Anon. Vol. 1 Publ by IEEE, 1993. p. 67-72.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rudas, I, Pereszlenyi, F, Tar, J & Bito, JF 1993, General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization. in Anon (ed.), IECON Proceedings (Industrial Electronics Conference). vol. 1, Publ by IEEE, pp. 67-72, Proceedings of the 19th International Conference on Industrial Electronics, Control and Instrumentation, Maui, Hawaii, USA, 11/15/93.
Rudas I, Pereszlenyi F, Tar J, Bito JF. General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization. In Anon, editor, IECON Proceedings (Industrial Electronics Conference). Vol. 1. Publ by IEEE. 1993. p. 67-72
Rudas, I. ; Pereszlenyi, F. ; Tar, J. ; Bito, J. F. / General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization. IECON Proceedings (Industrial Electronics Conference). editor / Anon. Vol. 1 Publ by IEEE, 1993. pp. 67-72
@inproceedings{ef5d468faf514f3d88530f89078bdd99,
title = "General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization",
abstract = "Recent development of adaptive and robust controllers for non-linear coupled systems is seriously inspired by new possibilities of hardware-realization. It concerns each kinds of artificial intelligence: the 'classical' knowledge-based systems (KBSs), the more modern artificial neural network- (ANN) and fuzzy sets- (FS) based formulations. While common and general mathematical roots of these approaches became transparent, in each concrete case complexity of the appropriate control rules, the number of the ANN processor units or fuzzy sets, complexity of the fuzzy rules of a satisfactory controller essentially depend on the generalized coordinates used for describing the system to be controlled. In this paper it is pointed out, that in a wide class of control tasks relatively simple kids of coordinate transformation can be applied to achieve effective simplification of the control rules and reduction of the number of the necessary processor units or fuzzy sets. As an example, dynamics of robot arms are considered, where the appropriate coordinate transformation is approximate diagonalization of the inertia matrix of the robot arms. It serves as a satisfactory basis for working out a general method and principles for dealing with the quadratic coupling not eliminated by the diagonalization. It shows strong robustness with respect to the unknown inertia of the manipulated body and easily can be realized by ANN or fuzzy rule-based solutions of very limited number of processor units or fuzzy rules. The method is demonstrated by results of numerical simulation for a concrete physical system. Effects of imprecise modeling of the mass of the work-piece carried by the robot are considered, too.",
author = "I. Rudas and F. Pereszlenyi and J. Tar and Bito, {J. F.}",
year = "1993",
language = "English",
isbn = "0780308913",
volume = "1",
pages = "67--72",
editor = "Anon",
booktitle = "IECON Proceedings (Industrial Electronics Conference)",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - General principles for design of robust non-linear controllers for AI-, ANN- or fuzzy approach-based realization

AU - Rudas, I.

AU - Pereszlenyi, F.

AU - Tar, J.

AU - Bito, J. F.

PY - 1993

Y1 - 1993

N2 - Recent development of adaptive and robust controllers for non-linear coupled systems is seriously inspired by new possibilities of hardware-realization. It concerns each kinds of artificial intelligence: the 'classical' knowledge-based systems (KBSs), the more modern artificial neural network- (ANN) and fuzzy sets- (FS) based formulations. While common and general mathematical roots of these approaches became transparent, in each concrete case complexity of the appropriate control rules, the number of the ANN processor units or fuzzy sets, complexity of the fuzzy rules of a satisfactory controller essentially depend on the generalized coordinates used for describing the system to be controlled. In this paper it is pointed out, that in a wide class of control tasks relatively simple kids of coordinate transformation can be applied to achieve effective simplification of the control rules and reduction of the number of the necessary processor units or fuzzy sets. As an example, dynamics of robot arms are considered, where the appropriate coordinate transformation is approximate diagonalization of the inertia matrix of the robot arms. It serves as a satisfactory basis for working out a general method and principles for dealing with the quadratic coupling not eliminated by the diagonalization. It shows strong robustness with respect to the unknown inertia of the manipulated body and easily can be realized by ANN or fuzzy rule-based solutions of very limited number of processor units or fuzzy rules. The method is demonstrated by results of numerical simulation for a concrete physical system. Effects of imprecise modeling of the mass of the work-piece carried by the robot are considered, too.

AB - Recent development of adaptive and robust controllers for non-linear coupled systems is seriously inspired by new possibilities of hardware-realization. It concerns each kinds of artificial intelligence: the 'classical' knowledge-based systems (KBSs), the more modern artificial neural network- (ANN) and fuzzy sets- (FS) based formulations. While common and general mathematical roots of these approaches became transparent, in each concrete case complexity of the appropriate control rules, the number of the ANN processor units or fuzzy sets, complexity of the fuzzy rules of a satisfactory controller essentially depend on the generalized coordinates used for describing the system to be controlled. In this paper it is pointed out, that in a wide class of control tasks relatively simple kids of coordinate transformation can be applied to achieve effective simplification of the control rules and reduction of the number of the necessary processor units or fuzzy sets. As an example, dynamics of robot arms are considered, where the appropriate coordinate transformation is approximate diagonalization of the inertia matrix of the robot arms. It serves as a satisfactory basis for working out a general method and principles for dealing with the quadratic coupling not eliminated by the diagonalization. It shows strong robustness with respect to the unknown inertia of the manipulated body and easily can be realized by ANN or fuzzy rule-based solutions of very limited number of processor units or fuzzy rules. The method is demonstrated by results of numerical simulation for a concrete physical system. Effects of imprecise modeling of the mass of the work-piece carried by the robot are considered, too.

UR - http://www.scopus.com/inward/record.url?scp=0027838855&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027838855&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0027838855

SN - 0780308913

VL - 1

SP - 67

EP - 72

BT - IECON Proceedings (Industrial Electronics Conference)

A2 - Anon, null

PB - Publ by IEEE

ER -