面向后件集的模糊推理机制: 用于区间型type-2模糊逻辑系统

目前,大多数模糊推理都是利用t-范数和t-余范数或其改进形式对连接词进行建模,这些模型不能将模糊规则中前件集与后件集之间的相关性信息引入到模糊推理过程,这会丢失蕴含在规则中的一些信息甚至导致推理结果与实际经验严重不符.为解决此问题,本文首先引入模糊集合面向对象变换的概念,并将其推广,建立了合成type-2模糊集合模型.基于此模型,针对区间型type-2模糊逻辑系统,提出一种面向后件集的模糊推理机制,该机制能将前件集与后件集的相关性信息(包括清晰数和模糊数两种情形)引入到模糊推理过程.仿真结果表明,该方法能捕获到模糊规则中更多的不确定性信息,并为模糊逻辑系统的设计提供更大的自由度.

Consequent-oriented fuzzy inference: for interval type-2 fuzzy logic systems

Most fuzzy inference methods usually use some kind of t-norm and t-conorm, or their improved forms, to model the connectives. Unfortunately, those models are unable to bring the relationship between antecedents and consequents into the process of fuzzy inference. This may lose some information implied in the rule and even leads to the inference results inconsistent with practical experiences. To solve this problem, the concept of object-oriented transform on fuzzy sets is introduced and extended to establish a compound type-2 fuzzy set model, by which a consequent-oriented fuzzy inference mechanism is developed. The mechanism can introduce the relationship between antecedents and consequents, including crisp numbers and fuzzy number, to the process of fuzzy inference. Simulation results indicate that the proposed method can capture more information about rule uncertainties, and provide more choices in designing a fuzzy logic system.