摘要
本文在带语言变量和可能性分布方程的模糊逻辑作为模糊知识表示的基础上,提出了一种有效的反问推理技术——推理树的生成与评估算法,这种算法基于扎德(L.A.Zadeh)的似然推理理论,并且该算法没有回溯、不会导致无穷循环。另外,由于使用了基于语言变量的软匹配方法,它有效地处理了模糊知识之间的匹配,从而使得模糊推理有效性和适用性大为提高。
An efficient backward inference technique is given by the generation andestimation of inference tree in this paper,and the fuzzy logic with linguistic va(?)iablesand possibility distributions is considered as a basic fuzzy knowledge representation.There are no backtracking and dead cycling in this algorithm,and the effectivenessand availiability can be greatly improved by using the matching among the fuzzyknowledges.
出处
《计算机应用与软件》
CSCD
1991年第6期49-56,共8页
Computer Applications and Software
