首页 | 本学科首页   官方微博 | 高级检索  
     检索      

集成节约覆盖集理论和模糊概率推理的方法
引用本文:钟毓宁,杨叔子,师汉民,史铁林.集成节约覆盖集理论和模糊概率推理的方法[J].华中科技大学学报(自然科学版),1991(Z2).
作者姓名:钟毓宁  杨叔子  师汉民  史铁林
作者单位:华中理工大学机械工程一系 (钟毓宁,杨叔子,师汉民),华中理工大学机械工程一系(史铁林)
基金项目:国家自然科学基金资助项目
摘    要:本文运用Bayes定理分析了各种故障发生的先验概率对在已知征兆下判断故障的准确性的影响情况;论述了使用模糊概率描述征兆、故障、征兆与故障之间的映射关系等的方法;在将模糊概率和节约覆盖集理论集成的基础上,提出了基于浅知识的新的诊断推理方法.该方法在一定程度上减小了获取有关概率知识的难度,并对概率值的定义误差有一定的容错能力.

关 键 词:人工智能  诊断问题求解  节约覆盖模型  模糊概率推理

A New Approach to Integrating the Parsimonious Covering Model and the Fuzzy Probabilistic Inference
Zhong Yuning Yang Shuzi Shi Hanmin Shi Tielin.A New Approach to Integrating the Parsimonious Covering Model and the Fuzzy Probabilistic Inference[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,1991(Z2).
Authors:Zhong Yuning Yang Shuzi Shi Hanmin Shi Tielin
Institution:Zhong Yuning Yang Shuzi Shi Hanmin Shi Tielin
Abstract:The parsimonious covering model is a main model for abductive rea-soning. Integrating the model and the probabilistic inference is a new approach to combine the symbol inference with the numerical probabilistic calculation. In this approach, all probabilities are treated as exact values in the diagnostic problem-solving. For many diagnostic cases, it is often difficult to evaluate these probabilities. Therefore, it is proposed that fuzzy probabilities be used. All abnormalities, symptoms and causalities are described by fuzzy probabilities. With the fuzzy probabilistic information available, the relative likelihood measures of possible diagnostic solutions of given symptoms can be calculated. The search for an optimum solution for concurrent diagnostic problem-solving is investigated.
Keywords:Artificial intelligence  Diagnostic problem-solving  Parsimoni- ous covering model  Fuzzy probabilistic inference  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号