An efficient identification scheme for nonlinear polynomial NARX model

Yu Cheng, Miao Yu, Lan Wang, Jinglu Hu

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

Abstract

Nonlinear polynomial NARX model identification often faces the problem of huge pool of candidate terms, which makes the evolutionary optimization based identification algorithm work with low efficiency. This paper proposes an efficient identification scheme with pre-processing to reduce the searching space effectively. Both the input selection and term selection are implemented to truncate the candidate pool with the help of correlation based orthogonal forward selection (COFS) algorithm and simplified orthogonal least square (OLS) algorithm, respectively. Then multi-objective evolutionary algorithm (MOEA) is used to identify the polynomial model in a relative small searching space.

Original languageEnglish
Title of host publicationProceedings of the 16th International Symposium on Artificial Life and Robotics, AROB 16th'11
Pages499-502
Number of pages4
Publication statusPublished - 2011 Dec 1
Event16th International Symposium on Artificial Life and Robotics, AROB '11 - Beppu, Oita, Japan
Duration: 2011 Jan 272011 Jan 29

Publication series

NameProceedings of the 16th International Symposium on Artificial Life and Robotics, AROB 16th'11

Conference

Conference16th International Symposium on Artificial Life and Robotics, AROB '11
CountryJapan
CityBeppu, Oita
Period11/1/2711/1/29

Keywords

  • Efficient
  • Input selection
  • Nonlinear polynomial model identification
  • Term selection

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Human-Computer Interaction

Fingerprint Dive into the research topics of 'An efficient identification scheme for nonlinear polynomial NARX model'. Together they form a unique fingerprint.

Cite this