摘要
通过将时态信息引入因果模型而构成时态因果诊断模型TCDM(TemporalCauseDiagnosticModel)。与此同时,给出了不确定性、子骨架以及关于TCDM的故障因果演变过程的概念,并通过对TCDM拓扑结构的分析提出了“分解-合成”原理。基于上述原理,提出了关于TCDM的“分解-合成”算子以及分解TCDM的策略。利用“分解-合成”算子可有效地分解或合成TCDM“子骨架”。而分解策略可将TCDM分解成若干相互关联的子模型,这些子模型有较好的结构。便于故障分离及分布式处理。还详细说明了因果模型分解、诊断问题求解、各子骨架部分解的合成等方法。
A temporal cause diagnostic model (TCDM) is developed by using temporal information in cause model. The concept of indeterminateness,sub-fromwork and fault causal developing process about the TCDM are defined. As a result of the topic structure analysing on TCDM, the principle of “splitting-synthesis” is presented. Based on these result, the operator of “splitting-synthesis” and the splitting stratagem of the TCDM is designed. Useing of the operator it is very effective to split or to synthes sub-fromwork. The splitting stratagem of the TCDM can be used to split up cause model into some correlative sub-model that have good constructure, is easily seperated and distributed when fail. Causal model splitting,diagnostic problem solving and part solution synthesizing on every sub-model are explained in detail.
出处
《石油化工高等学校学报》
EI
CAS
1998年第2期69-73,共5页
Journal of Petrochemical Universities
关键词
人工智能
故障诊断
因果模型
时态知识
Artificial intelligence
Fault diagnostic
Cause model
Temporal knowledge