vague命题逻辑与基于线性变换的近似推理  

Vague propositional logic and linear transformate based approximation inference

作　　者：冯志强[1,2] 柳存根[3] 

机构地区：[1]上海交通大学船舶海洋与建筑工程学院,上海200240 [2]江苏省船舶工程技术中心,江苏南通226010 [3]上海交通大学海洋工程国家重点实验室,上海200030

出　　处：《哈尔滨工程大学学报》2010年第9期1222-1227,共6页Journal of Harbin Engineering University

基　　金：国家重点科技攻关资助项目(教技函[2005]120)

摘　　要：针对vague命题与近似推理问题,研究了不同论域上的基本命题运算与推理表达.由集合的扩张原理,引入vague命题上的基于连续三角模的广义合取/析取以及S-蕴涵与R-蕴涵运算.依据合成运算的定义给出vague线性变换的概念,提出基于vague线性变换的近似推理方法,进而得出单规则、多规则以及复合前提多规则的推理形式的表达.通过命题运算可实现多论域的各种vague环境下信息的聚合.与其他方法相比较,基于线性变换的推理方法具有操作简单、可选择性强等特点.Basic operations of vague propositions and approximation inference rules were discussed as they apply over multi-universes.By the extension principles of collections,generalized conjunctions and disjunctions,S-implications and R-implications over vague propositions were introduced via fuzzy logical operators.As per the definition of compositional operations for vague relations,a notion of a vague linear transformation as well as an approach to approximation inference rules were given.Furthermore,the authors derived expressions of inference rules based on a single rule,multi-rules and rules with composite antecedents.By means of operations on propositions,information aggregation was implemented in vague environments.As compared with other methods,inferences based on linear transformations were found to be simpler and more selectable.

关 键 词：扩张原理 vague命题逻辑 线性变换 近似推理方法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]