基于粗糙集理论的飞机供电系统诊断规则提取

 为有效解决飞机供电系统故障诊断专家系统在知识获取方面的“瓶颈”，在对某型号飞机供电系统进行故障模式分析的基础上，利用粗糙集理论不需要提供待求解问题所需处理的数据集合之外的任何先验信息的特点，提出了一种诊断规则的自动获取方法。在运用粗糙集理论对故障样本集进行属性约简和值约简的过程中，针对“不确定测量状态”提出了改进的约简规则。通过在交流一次配电子系统上的验证分析表明，该方法具有较强的容错能力，可有效地约简知识，自动获取规则，且规则具有正确的逻辑意义。该方法为飞机供电系统故障诊断专家系统由理论验证向实际应用的转化提供了可行性。

Fault Diagnosis in Aircraft Power System Based on Rough Sets

For the purpose to resolve the bottleneck problem of obtaining knowledge of aircraft power system expert system,based on the fault pattern analysis of aircraft power system,according to the advantage that rough sets can effectively handle data sets without any pre-validated information,a method used to obtain diagnosis rule automatically is proposed.In the process of attribute reduction and value reduction,an improved method is presented to handle the unknown measuring state.The example in AC power distribution sub-system shows that the method is quite efficient,fault-tolerant,and able to obtain rules automatically which is correct in logic.This method provides the feasibility that the expert system in aircraft power system can be converted from theory to practice.