基于故障树的知识存储及诊断系统  

Fault tree-based knowledge storage and diagnosis system

作　　者：陈文钦[1] 王强[2] 李波[2] 

机构地区：[1]贵州大学计算机科学与信息学院,贵州贵阳550025 [2]电子科技大学空天科学技术研究院,四川成都610054

出　　处：《广西大学学报（自然科学版）》2010年第2期324-329,共6页Journal of Guangxi University（Natural Science Edition）

基　　金：四川省应用基础研究(2008JY0060)

摘　　要：分析半导体制造企业基于故障树的诊断系统的现状,针对目前传统故障树在关系型数据库的存储方式的不合理性,提出了新的在数据库中的分级存储方式,并基于此建立了新型专家系统。介绍基于故障树的知识表述方法,提出一种基于故障树的知识在关系型数据库中的分级存储方式并介绍基于混合推理的故障识别诊断方法;讨论实现诊断系统的关键技术,如数据存储的类实现方式和B/S三层架构模式等。经成都某半导体芯片封装测试工厂实际使用,所开发系统方便简捷,能有效提高故障诊断的效率和可靠性。The current situation of fault diagnosis system based on fault tree theory in semiconductor manufacturing was analyzed. According to the disadvantage of the traditional fault tree storage meth- od in the relational database, a new storage method was designed and a new diagnosis system was built based on this new storage form. Knowledge representation based on fault tree was introduced and the new storage method in relational database was provided firstly. The fault diagnosis based on the hybrid reason was introduced. The key technologies to implement the expert system were introduced, which are the usage of classes to define data storage methods and three-tier Brows/Server mode. The actual application in a chipset assembly and test factory at Chengdu demonstrates that the method and diagnosis system can be conveniently used and improve the efficiency and reliability of fault diagnosis.

关 键 词：故障树 故障诊断 知识存储 数据库 诊断系统 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]