摘要
不确定性推理问题是专家系统需要解决的基本问题之一,已有的不确定性推理方法在规则的应用要求、加权推理及规则参数获取方法方面具有局限性。该文在分析、归纳、总结不确定性推理相关问题的基础上,提出了基于充分性参数和必要性参数的不确定性推理方法。该方法能够表示具有复杂不确定性加权关系的诊断规则,并且能够根据规则的约束条件,通过数值计算获取规则的不确定性相关参数。该方法在已开发的诊断型专家系统中得到了应用,经验表明该方法是行之有效的。
Uncertain reasoning is one of the critical issues in expert systems. Current uncertainity reasoning methods have many restrictions, such as rule conditions, weight selection, and rule parameter acquisition. Analysis of these uncertainity reasoning issues led to the development of an uncertainity reasoning method with dual factors. The method includes diagnostic rules which include complicated relationships between the conditions. The factors related to the diagnostic rules uncertainties can be calculated based on the conditions in the rules. Application of the method in expert diagnostic systems verified the effectiveness of the uncertainity reasoning method.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第8期1397-1400,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家杰出青年科学基金资助项目(50425516)
教育部"跨世纪优秀人才培养计划"基金资助项目
关键词
故障诊断
不确定性推理
专家系统
uncertainity reasoning
expert system
fault diagnosis
