Fuzzy Inference Systems Composed of Double-Input Rule Modules for Obstacle Avoidance Problems

The purpose of self-tuning algorithm for fuzzy inference system is to construct automatically fuzzy inference rules from learning data based on the steepest descend method. Obvious drawbacks of the method are its large computational complexity and getting stuck in a shallow local minimum. Further, i...

Full description

Saved in:
Bibliographic Details
Published inIAENG international journal of computer science Vol. 41; no. 4; pp. 222 - 230
Main Authors Miyajima, Hirofumi, Kawai, Takehiro, Shigei, Noritaka, Miyajima, Hiromi
Format Journal Article
LanguageEnglish
Published 01.12.2014
Subjects
Computer simulation
Fuzzy
Fuzzy logic
Fuzzy set theory
Inference
Learning
Mathematical models
Modules
Online AccessGet full text

Cover

More Information
Summary:The purpose of self-tuning algorithm for fuzzy inference system is to construct automatically fuzzy inference rules from learning data based on the steepest descend method. Obvious drawbacks of the method are its large computational complexity and getting stuck in a shallow local minimum. Further, it is difficult to apply for the conventional method to the problem with a large number of variables. In order to overcome them, the SIRMs (Single-Input Rule Modules) and DIRMs (Double-Input Rule Modules) models have been proposed. In some numerical simulations, it is shown that there exists the difference of the ability between DIRMs and SIRMs models. In this paper, we will apply DIRMs and SIRMs models to the control problem of obstacle avoidance. As a result, it is shown that DIRMs model is more effective than SIRMs model in this problem. Further, we propose a learning method to reduce the number of modules of DIRMs model and show the effectiveness in numerical simulations.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:1819-656X
1819-9224