RUS  ENG
Full version
JOURNALS // Problemy Upravleniya // Archive

Probl. Upr., 2009 Issue 4, Pages 15–23 (Mi pu81)

This article is cited in 10 papers

Analysis and synthesis of control systems

Identification of fuzzy systems: methods and algorithms

I. A. Hodashinsky

Tomsk State University of Control Systems and Radioelectronics

Abstract: The paper considers three basic phases of fuzzy systems construction: expert evaluation, structure identification, parameter estimation. Expert evaluation includes: selection of fuzzy model type; choice of $t$-normal functions to set the fuzzy logic operations; choice of a fuzzy logic inference. For structure identification the fuzzy clustering method and iterative algorithm are offered. For parameters optimization the following methods have been chosen: genetic algorithm, ant colony algorithm, particle swarm optimization, simulated annealing.

Keywords: fuzzy system identification, metaheuristics, simulated annealing, genetic algorithm, ant colony algorithm, particle swarm techniques.

UDC: 004.82



© Steklov Math. Inst. of RAS, 2024