论文部分内容阅读
首先提出了一种基于实例的混合推理总体结构。在一级实例库和二级实例库概念的基础上，定义了实例的结构及组织。并设计了三级推理机制，即以故障代码为匹配目标、一级实例库为匹配空间进行的一级实例推理，以模糊化的初始故障征兆为匹配目标的二级实例推理。最后给出了故障征兆模糊处理以及实例匹配的方法。
First, an example based hybrid inference general structure is proposed. Based on the concepts of level one and level two instance libraries, the structure and organization of an instance is defined. And the three-level reasoning mechanism is designed, that is, the first level instance reasoning based on the fault code as the matching target and the first level instance space as the matching space, and the second-level instance reasoning using the fuzzy initial fault symptom as the matching target. Finally, the method of fuzzy treatment of fault symptom and instance matching is given.