Hits:

Indexed by:期刊论文

Date of Publication:2008-11-01

Journal:ASIAN JOURNAL OF CONTROL

Included Journals:SCIE、EI、Scopus

Volume:10

Issue:6

Page Number:687-697

ISSN No.:1561-8625

Key Words:Fuzzy modeling; fuzzy clustering; rule reduction; objective function

Abstract:A systematic fuzzy approach considering both accuracy and interpretability is developed in the paper. First, a fuzzy modeling method based on a new objective function is proposed. The proposed method can deal with the problem where the input variables have an affect on the input space of the fuzzy system while the output variables do not exert any influence on input space of fuzzy system. Then rule reduction is performed to obtain the model structure of the fuzzy system by QR decomposition of the fuzzy reference matrix. According to analysis of the rank loss of the matrix, the important rules and unimportant rules can be confirmed in this paper. Simulation results demonstrate that the proposed approach can be used to build fuzzy models of nonlinear systems.