摘要： 针对现代流程工业的特点，利用符号有向图模型(SDG)对化工生产工艺流程进行建模，运用SDG的推理能力，将其因果图转化为产生式规则，获得对整个生产流程的规则描述。根据SDG模型推理结果的特点提出一种可行的故障诊断专家系统设计方案，将得到的规则存入专家系统知识库，进行基于征兆的前向推理故障诊断。应用结果表明该种方法的有效性与可行性。

关键词: 符号有向图, 产生式规则, 知识库

Abstract: In view of the characteristics of modern chemical process flow, this paper models the process flow of chemical production by utilizing Signed Directed Graph(SDG) model, and obtains causality diagram that describes the mechanism of process flow. By using the reasoning ability of SDG, the causality diagram is transformed into production rule. And based on the reasoning features of SDG model, the paper puts forward a practicable design scheme for diagnosis expert system to store the obtained rules into knowledge base and conduct fault diagnosis by means of forward reasoning. The method has been proved effective and feasible in practical application.

Key words: Signed Directed Graph(SDG), production rule, knowledge base

中图分类号: 

• TP182