摘要
提出了一种基于粗糙神经网络的航空电子系统故障诊断方法。该方法运用粗糙集理论进行知识规则的提取,用提取的规则对航空电子系统输出数据进行分类,再用神经网络对每一类进行辨识,最后用粗糙集分类结果与各个神经网络输出进行综合得到系统一步预测结果,将该预测结果与航空电子系统实际输出进行阀值比较,进而进行故障诊断。实际应用表明该方法可行并能够有效地检测系统故障,故障诊断率高。
The paper presents a method of system fault diagnosis based on rough set neural network. Firstly, it discover knowledge by using rough set theory. Secondly, the method forecast the system output by combined discovered knowledge and several neural network. The method diagnose system fault by analyzed the forecast output and system real-time output. The application shows that the method is feasible and availability.
出处
《火力与指挥控制》
CSCD
北大核心
2006年第10期48-50,共3页
Fire Control & Command Control
