Data-driven construction of sugeno controllers: Analytical aspects and new numerical methods

Authors Josef Haslinger
Ulrich Bodenhofer
Martin Burger
Title Data-driven construction of sugeno controllers: Analytical aspects and new numerical methods
Booktitle Proc. Joint 9th IFSA World Congress and 20th NAFIPS Int. Conf. (IFSA-NAFIPS 2001)
Type in proceedings
Address Vancouver, BC, Canada
ISBN 0-7803-7079-1
Month July
Year 2001
Pages 239-244
SCCH ID# 115
Abstract

In this paper, the general problem of identifying the parameters describing Sugeno controllers from example data is addressed. It is shown that this approximation is ill-posed, which necessitates the use of regularization methods. Based upon this discovery, an efficient numerical algorithm for solving the general regularized approximation problem is proposed, followed by numerical examples with real-world data.