关于智能诊断中模型的构建和应用

在基于模型的诊断中,模型起着关键性的作用,模型的质量直接影响诊断的准确度和诊断的效率.因此诊断模型也成为人工智能研究青普遍关注的热点问题。长文对基于模型的诊断方法和该领域的最新研究成果做了全面的综述,重点介绍了构建诊断模型方面的研究成果,其中包括分级结构抽象方法、基于状态转换的历程重构方法、依赖于任务的定量抽象方法以及利用过程代数的描述方法等。此外,本文还对诊断模型的构建的研究提出了自己的观点和建议。

Construction and Application of the Models in Model-Based Diagnoses

In model based diagnoses,the models act as very important roles,which qualities effect the precision and effi- ciency of diagnosis so much that diagnosis model research becomes a hot problem for the AI researchers.Model-based diagnosis methods and some new research results in the field are completely summarized.Especially,the research results about models construction in diagnosis are mainly introduced,which include Hierarchical model-based diagnosisbased on structural abstraction,History reconstruction based on state transitions,Task-dependent qualitative domain ahstrac- tion and Model-based diagnosis based on process algebras,etc.Furthermore,some new ideas about the diagnosis model construction are proposed.