1/2010 - 17 |
Elite Based Multiobjective Genetic Programming in Nonlinear Systems IdentificationPATELLI, A. , FERARIU, L. |
View the paper record and citations in |
Click to see author's profile in SCOPUS, IEEE Xplore, Web of Science |
Download PDF (427 KB) | Citation | Downloads: 1,626 | Views: 5,599 |
Author keywords
evolutionary algorithms, genetic programming, multiobjective optimization, nonlinear system identification
References keywords
programming(7), genetic(6), evolutionary(6), systems(5)
Blue keywords are present in both the references section and the paper title.
About this article
Date of Publication: 2010-02-27
Volume 10, Issue 1, Year 2010, On page(s): 94 - 99
ISSN: 1582-7445, e-ISSN: 1844-7600
Digital Object Identifier: 10.4316/AECE.2010.01017
Web of Science Accession Number: 000275458900017
SCOPUS ID: 77954683610
Abstract
The nonlinear systems identification method described in the paper is based on genetic programming, a robust tool, able to ensure the simultaneous selection of model structure and parameters. The assessment of potential solutions is done via a multiobjective approach, making use of both accuracy and parsimony criteria, in order to encourage the selection of accurate and compact models, characterized by expected good generalization capabilities. The evolutionary process is implemented from an elitist standpoint, and upgraded by means of two original contributions, namely an adaptive niching mechanism and an elite clustering procedure. The authors have also suggested a set of enhancements to aid the genetic operators in effectively exploring the space of possible model structures. In symbiosis with the customized genetic operators, a QR local optimization procedure was integrated within the algorithm. It exploits the nonlinear, linear in parameter form that the working models are generated in, for providing a faster parameter computation. The performances of the proposed methodology were revealed on two applications, of different complexity levels: the identification of a simulated nonlinear system and the identification of an industrial plant. |
References | | | Cited By |
Web of Science® Times Cited: 2 [View]
View record in Web of Science® [View]
View Related Records® [View]
Updated 2 days, 18 hours ago
SCOPUS® Times Cited: 4
View record in SCOPUS® [Free preview]
View citations in SCOPUS® [Free preview]
[1] Parse-matrix evolution for symbolic regression, Luo, Changtong, Zhang, Shao-Liang, Engineering Applications of Artificial Intelligence, ISSN 0952-1976, Issue 6, Volume 25, 2012.
Digital Object Identifier: 10.1016/j.engappai.2012.05.015 [CrossRef]
[2] FPGA-Based Embedded System Architecture for Micro-Genetic Algorithms Applied to Parameters Optimization in Motion Control, JAEN-CUELLAR, A. Y., MORALES-VELAZQUEZ, L., ROMERO-TRONCOSO, R., OSORNIO-RIOS, R. A., Advances in Electrical and Computer Engineering, ISSN 1582-7445, Issue 1, Volume 15, 2015.
Digital Object Identifier: 10.4316/AECE.2015.01004 [CrossRef] [Full text]
[3] Identification and prediction using symbolic regression alpha-beta, Torres-Treviño, L. M., Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation, ISBN 9781450328814, 2014.
Digital Object Identifier: 10.1145/2598394.2609859 [CrossRef]
Disclaimer: All information displayed above was retrieved by using remote connections to respective databases. For the best user experience, we update all data by using background processes, and use caches in order to reduce the load on the servers we retrieve the information from. As we have no control on the availability of the database servers and sometimes the Internet connectivity may be affected, we do not guarantee the information is correct or complete. For the most accurate data, please always consult the database sites directly. Some external links require authentication or an institutional subscription.
Web of Science® is a registered trademark of Clarivate Analytics, Scopus® is a registered trademark of Elsevier B.V., other product names, company names, brand names, trademarks and logos are the property of their respective owners.
Faculty of Electrical Engineering and Computer Science
Stefan cel Mare University of Suceava, Romania
All rights reserved: Advances in Electrical and Computer Engineering is a registered trademark of the Stefan cel Mare University of Suceava. No part of this publication may be reproduced, stored in a retrieval system, photocopied, recorded or archived, without the written permission from the Editor. When authors submit their papers for publication, they agree that the copyright for their article be transferred to the Faculty of Electrical Engineering and Computer Science, Stefan cel Mare University of Suceava, Romania, if and only if the articles are accepted for publication. The copyright covers the exclusive rights to reproduce and distribute the article, including reprints and translations.
Permission for other use: The copyright owner's consent does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific written permission must be obtained from the Editor for such copying. Direct linking to files hosted on this website is strictly prohibited.
Disclaimer: Whilst every effort is made by the publishers and editorial board to see that no inaccurate or misleading data, opinions or statements appear in this journal, they wish to make it clear that all information and opinions formulated in the articles, as well as linguistic accuracy, are the sole responsibility of the author.