Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/90953
Type: Artigo de periódico
Title: Convex Relaxations With Guaranteed Convergence For Control Design Of Takagi-sugeno Fuzzy Systems [relaxações Convexas De Convergência Garantida Para O Projeto De Controladores Para Sistemas Nebulosos De Takagi-sugeno]
Author: Montagner V.F.
Oliveira R.C.L.F.
Peres P.L.D.
Abstract: This paper provides convex conditions with certificates of convergence for the design of state feedback controllers that quadratically stabilize and also ensure op- timal H2 and H∞ performances under quadratic sta- bility for Takagi-Sugeno continuous-time fuzzy systems. The proposed conditions are formulated as parameter- dependent linear matrix inequalities (LMIs) that have extra variables from Finsler's lemma and parameters belonging to the unit simplex. The fuzzy control law is written as a state feedback gain that is a homoge- neous polynomial of degree g, encompassing the parallel distributed compensator as a special case when g = 1. Algebraic properties of the system parameters and re- cent results of positive polynomials are used to construct LMI relaxations which, differently from most relaxations in the literature, asymptotically converge to a solution whenever such solution exists. Due to the degree of free- dom obtained with the extra variables, the conditions presented in the paper are an improvement over earlier results based on Pólya's theorem and can be viewed as an alternative to the use of techniques based on the relaxation of quadratic forms. The numerical efficiency in terms of precision and computational effort is demon- strated by means of comparisons with other methods from the literature.
Editor: 
Rights: aberto
Identifier DOI: 
Address: http://www.scopus.com/inward/record.url?eid=2-s2.0-77949931343&partnerID=40&md5=3cef0a74e548382a6736fc34a2a56df1
Date Issue: 2010
Appears in Collections:Unicamp - Artigos e Outros Documentos

Files in This Item:
File Description SizeFormat 
2-s2.0-77949931343.pdf233.91 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.