Lattice Structure of Binary-Equivalent Fuzzy Logical Functions under Max-Min Logic
Lattice Structure of Binary-Equivalent Fuzzy Logical Functions under Max-Min Logic
复制标题
最大-最小逻辑下二元等价模糊逻辑函数的格结构
DOI:
--
复制
发表时间:
2005
期刊:
影响因子:
--
通讯作者:
K. Hirota
中科院分区:
文献类型:
--
作者:
S. Yoshida;K. Hirota
Lattice structure of binary-equivalent fuzzy logical functions are analyzed under Max-Min fuzzy logic, Conventional studies of fuzzy flip-flops has shown that some different types of fuzzy flip-flops, whose binary logical characteristics are same, exist under Max-Min fuzzy logic. The fuzzy characteristics of these fuzzy flip-flops construct lattice structures (a Boolean lattice or a distributive lattice) under the partial order of ambiguity. This paper shows that a set of fuzzy extensions of a binary logical function constructs a distributive lattice for any binary logical function under Max-Min fuzzy logic. For analysis of a fuzzy logical function under Max-Min fuzzy logic, it is sufficient that the functions's value are examined for 0, 1/2, 1 of variables's values. (i.e, The characteristics of a fuzzy logical function are completely decided under the ternary logic using Max-Min operations.) When a binary logical function is extended to Max-Min logic, some of the functions's values can take a non-ambiguous value (0 or 1) or a ambiguous value (1/2). This difference derives the variations of fuzzy extensions of a binary logical funciton. If the variables' values are fixed, the variations of fuzzy extensions are at most 2. And the two variations are linearly ordered under both of the partial order of ambiguity. This is two-element Boolean lattice. If the variations of fuzzy extensions are decided independently for any variables' value, the structure of the fuzzy extensions contstruct the direct product of two-element Boolean lattice, i.e., n-element Boolean lattice. In some cases of fuzzy extensions, variations of function's value cannot be decided independently from other function's value. In this case, the structure of the extensions constructs non-Boolean distributive lattice.