一种诊断型专家系统的设计与实现

描述一种诊断型专家系统，它统属于基于规则的产生式系统，推理机是核心部分，分别采用了正向、反向推理技术、在不确定性知识处理上，应用确定性理论，概率统计理论和证据理论分别建立了不确定性推理模型。本系统采用关系模式来描述知识库，用数据库方法存储管理知识库，以改善知识库的性能，克服专家系统不能有效地利用现有知识库的局限。

An Approch to Design Diagnostic Exprert System

The system is a rule-based production system. Inference engine is the core part of expert system. We use forward chaining and backward chaining at the same time. In dealing with the uncertainty knowledge, we build up the uncertainty inference module based on certainty theory, evidence theory and probability theory. An approach to design knowledge base of expert system is used in the system . We describe a knowledge base by relation schema and use the database method to store and manage the knowledge base, to improve the property of knowledge base management , and to surmount the limitation of which an expert system can not use knowledge base effectively.