复杂航空电子装备故障诊断规则提取  

Extracting Diagnosis Rules of Complicated Avionics Under Condition of Small-Sample

作　　者：罗晓军[1] 文莹[2] 方甲永[2] 

机构地区：[1]桂林空军学院,广西桂林541003 [2]空军工程大学,陕西西安710038

出　　处：《计算机测量与控制》2012年第2期294-296,共3页Computer Measurement &Control

基　　金：国家部委基金资助项目

摘　　要：针对复杂航空电子装备故障样本不足的问题,为简化测试过程、提高故障诊断效率,提出了一种基于粗糙集的故障诊断规则提取方法;该方法构建了故障诊断决策信息系统,通过构造故障诊断决策信息系统的决策辨识矩阵和决策辨识函数,计算测试参数集的所有约简,从而提取故障诊断规则;以某型雷达系统为例,采用上述方法进行规则提取,将故障诊断所需测试参数从9个约简为3个;研究结果表明,该方法能在故障样本较少的条件下有效约简测试参数集,提取故障诊断规则,有利于简化测试过程,提高诊断效率。With the problem of lacking fault samples of complicated avionics,in order to simplify test process and improve diagnosis efficiency,a method for extracting diagnosis rules based on rough set was promoted.A fault diagnosis decision-making information system was constructed to calculate all reductions of test-parameter-set by building decision-making distinguishing matrix and decision-making distinguishing functions of the fault diagnosis information system,so that the fault diagnosis rules can be extracted.For example,the method for extracting diagnosis rules was used in the diagnosis of some radar system,with the result that 9 test parameters used for diagnosis were reduced to 3.The result of the study shows that the method can efficiently reduce the test-parameter-set and extract fault diagnosis rules under the condition of small-sample.And it is helpful for simplifying test process and improving the diagnosis ability.

关 键 词：小样本 粗糙集 故障诊断 复杂航空电子装备 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]