基于规则的机械故障诊断系统的验证模型与方法

VERIFICATION MODEL AND METHOD FOR THE RULE-BASED MECHANICAL EXCEPTION DIAGNOSIS SYSTEM

机械领域知识库是机械故障诊断专家系统的核心部分,知识库的异常会导致整个推理求解过程的不确定性,引起一系列的错误和冲突。将Petri网技术应用于机械故障诊断系统规则库的建模和验证,给出了规则库的有色Petri网的建模方法,对规则的使用次数控制、事实的保留等问题给出了处理技巧。基于有色Petri网得到了知识库各种异常现象的具体表现以及检测的方法。所给出的有色Petri网建模方法和分析技术适应于各个领域诊断系统的验证和分析,所提出的技术和方法具有普遍意义。

Mechanical domain knowledge base is a central component of the domain diagnosis expert system, so its abnormity, such as redundancy, and conflict, can affect the reasoning result. The mechanical knowledge is taken as backgrounds, and the Petri net technology is introduced for the knowledge analysis and verification. The methods modeling the knowledge base with the colored Petri net are presented, which include the save of the facts, the control of using the same rules and so on. Based on the colored Petri net, the methods are presented for detecting the abnormity of the mechanical domain knowledge base.