Title: Fuzzy system identification based on support vector regression and genetic algorithm

Authors: Wei Li, Yupu Yang, Zhong Yang, Changying Zhang

Addresses: Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China. ' Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China. ' Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China. ' Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China

Abstract: A new fuzzy identification approach using support vector regression (SVR) and genetic algorithm (GA) is presented in this paper. Firstly positive definite reference function is utilised to construct a qualified Mercer kernel for SVR. Then an improved GA is developed for parameters selection of SVR, in which the number of support vectors and regression accuracy are regarded simultaneously to guarantee the conciseness of the constructed fuzzy model. Finally, a set of TS fuzzy rules can be extracted from the SVR directly. Simulation results show that the resulting fuzzy model not only costs less fuzzy rules, but also possesses good generalisation ability.

Keywords: fuzzy system identification; positive definite reference function; support vector regression; SVR; genetic algorithms; GAs; fuzzy identification; simulation; fuzzy modelling.

DOI: 10.1504/IJMIC.2011.037829

International Journal of Modelling, Identification and Control, 2011 Vol.12 No.1/2, pp.50 - 55

Published online: 21 Mar 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article