摘要
狭义上,故障诊断包括两层含义即故障检测和故障定位,前者是指测试一个系统,判断其是否存在故障,而后者则是指在前者的基础上进一步定位故障点。广义上,故障诊断的内容,除了包含要判断系统有无故障以及故障产生后找到故障部位,还包含对故障特性的分析以及所采取的措施。
In this paper,fault diagnosis is studied as a classificatory problem,we constructed a universal fault model in order to grant the generality of the approach. The algorithm IHMCAP which combines symbolic and neural learning balances the uses of the dynamic information and the experimental knowledge in the system fault diagnosis. In the meantime,the incremental learning ability of the algorithm also can acquire highly diagnostic accuracy and capability of learning and diagnosis of the novel faults.
出处
《计算机科学》
CSCD
北大核心
1999年第1期85-87,84,共4页
Computer Science
