故障诊断专家系统的可拓知识表示和匹配研究

Study on Extension Knowledge Representation and Matching of Fault Diagnosis Expert System

摘要:

知识表示和匹配是设计专家系统的核心问题；首先通过引入基元理论,实现了产生式、语义网络、框架和案例的基元表示,建立了故障诊断专家系统的可拓知识表示模型,该模型包括激励基元、测量基元和结论基元3个部分；然后提出了先匹配激励基元再匹配测量基元的可拓知识匹配步骤,并构建了属性精确值和区间值混合的匹配度计算公式；最后以某型大气数据计算机的测试数据为例,与建立的可拓知识进行匹配,诊断结果与实际情况相符。

Abstract:

Knowledge representation and matching are key problems of expert system. By inducting basic-element theory, the basic-element representation of production, semantic network, frame and case are fulfilled, and extension knowledge representation model of diagnostic expert system is set up, which contains source, test and result three basic-element. The matching process needs two steps, first match source basic-element, second match test basic-element. At last, the test data of an air data computer is used to match the extension knowledge. The results are consistent with the fact.