摘要
试图为知识工程机器智能提供一种新的不确定性推理的框架.利用因素空间理论对推理句的因素机理进行了刻划,建立了推理渠道集结构,把推理描写成真值沿推理渠道的流动过程,在此基础上,对现有模糊推理的基础——推理关系作了严格的分析处理,澄清了其中的一些混乱,把模糊蕴涵的赋值公式从纯经验提到论证的高度.为了突破推理的静态研究局面,引入了兴奋度概念,建立了推理的动态分析模型.
The fuzzy reasoning models have been applied rapidly in several areas such as fuzzy control and pattern recognition. However there is lacking sufficient theoretical foundation still for its development. A new idea of treating inference process as the flow of truth values along inference channels is presented. By means of the factors space theory, a serious theory called truth-valued flow inference theory is obtained. By this way, the theory of inrference relation, which is the base of existent models of fuzzy rea- soning but has not been made into perfection, is retreated and reformed. Finally, a dy- namic analysis to this theory is given.
出处
《北京师范大学学报（自然科学版）》
CAS
CSCD
1989年第1期1-12,共12页
Journal of Beijing Normal University(Natural Science)
基金
国家自然科学基金
关键词
真值流推理
推理关系
人工智能
truth-valued-flows, inferences, inferences relations, exciting grade, net-inferences.
