GSI模糊推理模型与其他模糊推理模型的关系  

On the relationship between GSI and other methods of fuzzy reasoning

作　　者：谷敏强[1] 姜昆[2] 刘智斌[3] GU Minqiang;JIANG Kun;LIU Zhibin(Department of Mathematics,College of Science,Shantou University,515063,Shantou,Guangdong,China;Department of Natural Science,Shantou Polytechnic,515041,Shantou,Guangdong,China;College of Mathematics,Physics and Information Engineering,Zhejiang Normal University,324001,Jinhua,Zhejiang,China)

机构地区：[1]汕头大学理学院数学系,广东汕头515063 [2]汕头职业技术学院自然科学系,广东汕头515041 [3]浙江师范大学数理与信息工程学院,浙江金华324001

出　　处：《北京师范大学学报（自然科学版）》2018年第4期449-457,共9页Journal of Beijing Normal University(Natural Science)

基　　金：国家自然科学基金资助项目(11471202);广东高校国家级重点培育平台资助项目(2015KGJHZ014)

摘　　要：提出GSI(guaranteed similarity-degrees of fuzzy inference method)模糊推理模型的概念,考察该推理模型与其他典型的模糊推理模型之间的关系,证明如在计算公式中使用剩余型伴随对,输入前件为正规模糊集,常用的模糊推理算法如CRI算法、三I算法、基于∧-→合成的FRI算法都是GSI模糊推理算法,由Turksen提出的AARS算法(基于相似度的近似类比推理算法)是GSI模糊推理算法的近似.算例表明GSI模糊推理算法对系统输入比较敏感,具有良好的输入条件区分性能.GSI算法参数可调,具有广泛的适用性.The GSI fuzzy reasoning model is proposed and relationship between GSI and other methods of fuzzy reasoning is examined in the present work.The commonly used fuzzy reasoning such as CRI,triple I,and FRI based on∧-→composition are all similar to GSI,and Turksen’s AARS（approximate analogical reasoning schema）is a partial approximation of GSI.GSI is good to distinguish input conditions,and its weighted GSI model provides a wide range of adaptability and flexibility which would facilitate model optimization.

关 键 词：相似测度 蕴涵算子 GSI方法 

分 类 号：O142[理学—数学]