基于数值模拟的语言计算方法

An approach to computing with words based on numerical scale

为了描述模糊信息并建立模糊逻辑,研究了基于扩展原理和符号化模型的语言计算方法,提出基于数值模拟的对称分布语言变量表示及计算模型.该模型用特征值和标准方差二维数值反映语言的共性和特性.为了扩大运用范围,针对用任意三角模糊数表示的语言变量,提出基础语言集的寻找方法,以实现语言变量统一,完成模糊语言之间的逻辑运算.在此基础上给出基于数值模拟表示方法的语言计算规则,并证明其正确性.该方法将隶属度函数包含的信息模拟成数值,不仅保留了语言的模糊性信息,而且实现了符号化语言的便捷性,有助于模糊推理和模糊决策.

In order to describe fuzzy information and establish fuzzy logic,this paper studies an approach to computing with words based on extension principle and linguistic symbolic computational models.Numerical scale fuzzy linguistic representation and computational models are given.In this model,canonical characteristic value and standard deviation are used to reflect the common and special characters of words.In order to enlarge the range of application,an approach to building basic linguistic term sets is given to unify arbitrary linguistic labels to complete simple arithmetic operation on two-dimensional numbers.Computing with words can be done through computing with numerical values.The rule of computing with words is given and its correctness is proved.This method of transforming words into numerical values preserves the vagueness of words,and realizes the convenience of symbolic computational models.It is helpful for fuzzy reasoning and fuzzy decision making.