Earticle

초록

영어

This paper proposes a class of General Piecewise-Affine (GPWA) AutoRegressive eXogenous (GPWARX) models for nonlinear black-box identification. A GPWARX model is a weighted sum of GPWA basis functions. In n-dimensions, a GPWA basis function is defined over n+2 polyhedral regions, which are pairwise directly adjacent in the domain space. The geometrical structures of GPWA basis functions provide the GPWARX models with stronger representation capability and higher approximation efficiency in estimation of nonlinear systems than the other canonical PWA functions. Benchmark nonlinear identification examples are illustrated to show that GPWARX models have better estimation accuracy than HHARX and BPWARX models with the same number of parameters and using the same parameter identification algorithm.

목차

Abstract
 1. Introduction
 2. Preliminaries
 3. PWA Approximation Theorem
  3.1. PWA Models
  3.2 PWA Approximation Theorem
 4. GPWA Representation Model
  4.1 GPWA Basis Function
  4.2 GPWA Approximation
 5. Numerical Examples
 6. Conclusions
 References

키워드

저자

• Weiquan Sun [ School of Business Administration, Duquesne University, Pittsburgh, PA, USA ]
• Min Li [ School of Medicine, Indiana University ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Control and Automation

• 간기

  월간

• pISSN

  2005-4297

• 수록기간

  2008~2016

• 십진분류

  KDC 505 DDC 605