通过揭示模糊推理的一种CRI算法的逻辑原理,提出了模糊推理的一种新算法。并对单规则的CRI算法(或三I算法)的重心(或最大隶属度)去模糊法的有效性进行了研究。发现在单规则的CRI算法(或三I算法)的模糊推理中,当大前提的后件取对称模糊集时,采用重心去模糊法使得推理结果无效,而采用最大隶属度去模糊法可使得推理结果有效。
By opening out the logical principle of a type of CRI method in fuzzy reasoning, the paper present a new type of method, the validity of methods for barycenter(or maximum subjection degree)- removing fuzziness based on CRI methods (or Triple I method) under one rule, are discussed. The results are got as follows: in CRI methods (or Triple I) method under one rule, if the latter in major premise is taken as symmetrical fuzzy set, then using barycenter-removing fuzziness method is not effective to fuzzy reasoning; if the latter in major premise is taken as symmetrical fuzzy set, then using maximum subjection degree-removing fuzziness method is effective to fuzzy reasoning.