全蕴涵三I算法的推理结果研究  

Study on the Results of Triple I Method

作　　者：何映思[1] 全海金[2] 

机构地区：[1]西南大学计算机与信息科学学院,重庆400715 [2]西南大学数学与统计学院,重庆400715

出　　处：《计算机科学》2012年第10期248-250,共3页Computer Science

基　　金：西南大学博士基金项目(swu111051)资助

摘　　要：Zadeh提出的CRI算法是模糊推理中运用最广泛的算法,但其逻辑语义不太清楚,也不具有还原性。王国俊教授提出的全蕴涵三I算法具有还原性,并在一定程度上解决了模糊推理的逻辑基础不够完善的问题。现在,全蕴涵三I算法已经获得了普遍的接受,并有许多学者对该算法进行了多种形式的推广。但全蕴涵三I算法及其改进算法在实际控制中都还未见成功的运用。对全蕴涵三I算法的推理结果进行了全面分析,指出全蕴涵三I算法存在的问题及其不能应用于实际控制的原因。Zadeh＇s CRI method is applied in fuzzy reasoning wildly,but it has been criticized to be too complex and it is not consistent.To alleviate CRI method＇s drawback,Triple I Method was proposed by Professor Wang Guojun.Now,Triple I method is widely recognized as one of the most important fuzzy reasoning methods,and many existing fuzzy inference methods are based on it.Despite Triple I method＇s success in various research fields,it still has no effect on fuzzy control.This paper analyzed Triple I method.The results show that Triple I method has some drawback and it cannot be applied in control.

关 键 词：模糊推理 CRI算法 全蕴涵三I算法 蕴涵算子 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]