Group-based Evolutionary Swarm Intelligence for Recurrent Fuzzy Controller Design

Author

Juang, Chia-Feng ; Chung, I-Fang ; Chen, Shin-Kuan

Author_Institution

Nat. Chung Hsing Univ., Taichung

fYear

0

fDate

0-0 0

Firstpage

1710

Lastpage

1714

Abstract

Recurrent fuzzy controller design by the hybrid of multi-group genetic algorithm and particle swarm optimization (R-MGAPSO), is proposed in this paper. The recurrent fuzzy controller designed here is the Takagi-Sugeno-Kang (TSK)-type recurrent fuzzy network (TRFN). Both the number of fuzzy rules and parameters in TRFN are designed concurrently by R-MGAPSO. Evolution of population consists of three major operations: group enhancement by particle swarm optimization, variable-length individual crossover and mutation. To verify the performance of R-MGAPSO, control of a dynamic plant is simulated and compared with other genetic algorithms.

Keywords

control system synthesis; fuzzy control; fuzzy neural nets; genetic algorithms; learning (artificial intelligence); neurocontrollers; particle swarm optimisation; recurrent neural nets; Takagi-Sugeno-Kang type network; fuzzy rule; group-based evolutionary swarm intelligence; multigroup genetic algorithm; particle swarm optimization; recurrent fuzzy controller design; Algorithm design and analysis; Feedback loop; Fuzzy control; Fuzzy neural networks; Fuzzy systems; Genetic algorithms; Genetic mutations; Learning systems; Particle swarm optimization; Takagi-Sugeno-Kang model;

fLanguage

English

Publisher

ieee

Conference_Titel

Fuzzy Systems, 2006 IEEE International Conference on

Conference_Location

Vancouver, BC

Print_ISBN

0-7803-9488-7

Type

conf

DOI

10.1109/FUZZY.2006.1681936

Filename

1681936