摘要
针对航空雷达故障诊断知识不足的问题,研究了基于概念格的诊断规则提取方法;采集故障样本数据建立航空雷达故障诊断形式背景,设计了基于属性的增量式建格算法,并用于建立故障诊断形式背景的概念格;进而构造故障诊断形式背景的可辨识属性矩阵和决策辨识函数,计算测试参数集的所有约简,从而提取诊断规则;用历史故障记录验证诊断规则的有效性,诊断所需测试参数从9个约简为4个,诊断准确率达到83.3%;研究结果表明,该方法能分类处理故障样本,有效约简测试参数集,提取诊断规则,简化测试过程,提高诊断效率。
A method for extracting diagnosis rules based on concept lattice was promoted to solve the problem of lack of diagnosis knowl edge in airborne radar fault diagnosis. Fault samples were collected to build diagnosis formal context. An attributebased incremental algo rithm was designed to construct concept lattice of the diagnosis formal context. The decision distinguishing matrix and decision distinguishing function of the diagnosis formal context were formed to calculate all reductions of test parameters set. Diagnosis rules were deduced from such reductions. The history fault records validate that these diagnosis rules reduce test parameters needed for diagnosis from 9 to 4, and have a precision of 83.3% in diagnosis. Results show that the method can classify fault samples, reduce test parameters and extract diagnosis rules efficiently, which is useful to simplify test process and improve diagnosis efficiency.
出处
《计算机测量与控制》
北大核心
2013年第10期2612-2614,共3页
Computer Measurement &Control
关键词
概念格
属性约简
故障诊断
航空雷达
concept lattice attribute reduction fault diagnosis airborne radar
