Pseudo rule-driven model for a programmable T-norm

Abstract

-Rule-Driven processing has been proven to be a way of achieving high speed in fuzzy processing. Up to now, Rule-Driven architectures where designed to work with MIN or PROD as T-norm because they were the most commonly used in applications. This paper proposes a new Rule-Driven model valid for any T-norm (programmable T-norm) and any kind of membership functions (MF), provided the overlap factor is two, and they are a fuzzy partition. The architecture proposed can be implemented either by hardware or software.