基于D-S证据理论的模糊规则匹配方法

产生式规则是智能系统中大量采用的一种知识表示模式.规则常带有不确定性,规则前提中的变量值在很多情况下是一个模糊集,而当前事实中的变量值可能是模糊集、实数域上的区间值或者清晰数.为了统一处理这些情形,提出了基于证据理论,通过对两种贴近度进行加权合成计算规则匹配度的方法.该方法可根据变量值的分布特点自适应地调整每种贴近度的权重,从而使规则的匹配结果更为合理.

A Matching Approach of Fuzzy Rules Based on the D-S Evidence Theory

As a representation of knowledge,production rule is widely used in intelligent system.For rules with uncertainty,the variable in the promise of the rules is a fuzzy set in most of conditions,but the variable in the current fact is usually a fuzzy set,general number or section on the real number domain.To process the different variable,a new approach to calculate the matching degree was proposed by combine the two closeness degree distributed with different weights based on evidential theory.The weight of each closeness degree can be adjusted(adaptively) in terms of the distribution of variables,so the matching result is much more reasonable.