Dynamic system identification using a Type-2 Recurrent Fuzzy Neural Network

Author

Juang, Chia-Feng ; Lin, Yang-Yin ; Chung, I-Fang

Author_Institution

Dept. of Electr. Eng., Nat. Chung-Hsing Univ., Taichung, Taiwan

fYear

2009

Firstpage

768

Lastpage

772

Abstract

This paper proposes an interval type-2 recurrent fuzzy neural network (IT2RFNN) for dynamic system identification. The antecedent parts in each recurrent fuzzy rule in the IT2RFNN are interval type-2 fuzzy sets, and the consequent part is of the Takagi-Sugeno-Kang (TSK) type with interval weights. The recurrent structure in the T2RFNN enables it to handle dynamic system identification problems with a priori knowledge of system input and output delay numbers. A T2RFNN is constructed using concurrent structure and parameter learning. Simulations on dynamic system identification with clean and noisy outputs verify the performance of T2RFNN.

Keywords

fuzzy neural nets; fuzzy set theory; fuzzy systems; identification; learning (artificial intelligence); nonlinear dynamical systems; recurrent neural nets; IT2RFNN; Takagi-Sugeno-Kang fuzzy system; concurrent structure; delay number; interval type-2 fuzzy set; interval type-2 recurrent fuzzy neural network; interval weight; nonlinear dynamic system identification problem; parameter learning; recurrent fuzzy rule; Control systems; Fuzzy control; Fuzzy neural networks; Fuzzy sets; Fuzzy systems; Knowledge based systems; System identification; Takagi-Sugeno-Kang model; Uncertainty; Working environment noise;

fLanguage

English

Publisher

ieee

Conference_Titel

Asian Control Conference, 2009. ASCC 2009. 7th

Conference_Location

Hong Kong

Print_ISBN

978-89-956056-2-2

Electronic_ISBN

978-89-956056-9-1

Type

conf

Filename

5276371