针对对智能仪器进行故障维护需要耗费大量人力物力且效率较低的问题,提出了使用专家系统进行故障维护的方案.介绍了专家系统的结构和不确定性推理方法.以智能金属检测机为实例,建立了故障分析树并据此导出故障诊断规则树,利用可信度进行基于规则的正向推理,使用Ruby语言构建了故障诊断专家系统.使用20个故障样本对其进行测试,结果表明该专家系统能够正确诊断大多数故障,可供实际使用.%It takes much manpower and material resources to maintain intelligent instruments. To solve this problem,a solution of using expert system was proposed. The structure of expert system and a method of uncertainty reasoning was introduced. Faults of intelligent metal detector were studied and classified in order to draw the failure analysis trees and rule trees. Using certainty factor for rule-based forward reasoning,a fault diagnosis expert system was built up with Ruby language. The system was tested with 20 fault samples and the results show that the sytem is able to diagnose most of the faults and can be used for practical produciton.
展开▼
