A Tuning Methpd of Fuzzy Rules by Fuzzy SIngleton-Type Reasoning Method

摘要

In this paper, we propose a learning algorithm for tuning fuzzy rules under fuzzy singleton-type reasoning method. The main advantage of the method is that the antecedent parts, consequent parts and their weights can be learned at the same time without changing the form of the fuzzy rule table, which is different from the conventional learning algorithms. Furthermore, we illustrate the efficiency of the proposed method by several numerical examples.