Abstract: Aiming at the problem of current expert system about imprecise semantic representation of domain knowledge and low reasoning efficiency, an Ontology-based expert system framework is proposed. In accordance with rule structure, rule pattern Ontology and action Ontology are created. User reasoning conditions are transformed into sub-pattern Ontology, and then a reasoning algorithm of expert system based on similarity calculation based on name and instance as well as Clips reasoning mechanism is designed. An application example from high performance machining is given. Experimental results show that the system can obviously improve reasoning efficiency and semantic expression of reasoning conclusions.

Key words: expert system, rule base, reasoning

摘要： 针对当前专家系统领域知识语义表示不精确及推理效率低的问题，提出一种基于Ontology的专家系统框架。根据规则结构建立规则模式Ontology和动作Ontology，将用户推理条件作为模式子Ontology，设计出基于名称和实例的相似度计算及Clips推理机制的专家系统推理算法。结合高性能切削加工应用背景给出一个应用实例，应用结果表明，该系统能够明显提高推理效率和推理结果的语义表达。

关键词: 专家系统, 规则库, 推理

CLC Number: 

• TP391