基于信息熵属性约简的航空发动机故障诊断

针对不协调信息条件下的航空发动机故障诊断问题,研究了基于信息熵属性约简的故障诊断方法。首先定义了故障诊断信息系统来描述不协调故障样本数据,针对基本粗糙集模型分类能力不足的问题,引入变精度粗糙集模型处理不协调诊断信息系统;然后针对现有条件熵不能区分不确定性规则的缺陷,提出了变精度条件熵作为属性重要度的度量标准,设计了启发式属性约简算法,提取故障诊断规则。将该方法用于航空发动机故障诊断,验证了该方法可有效处理不协调信息,显著提高了航空发动机故障诊断的准确率。

Aero-engine fault diagnosis based on information entropy attribute reduction

A diagnosis method is proposed to deal with the inconsistent information in aero-engine fault diagnosis based on information entropy attribute reduction.Firstly,a diagnosis information system is defined to describe inconsistent information.A variable precision rough set model is introduced to dispose the sample data set and improve the classification quality of base rough set.Then variable precision conditional entropy is proposed to serve as the measurement of attribute significance and overcome the limitation of current conditional entropy in distinguishing uncertain rules.Based on the new attribute significance,a heuristic attribute reduction algorithm is designed to extract the diagnosis rules.This approach was used in the fault diagnosis of aero-engine.Results show that the developed approach can deal with inconsistent information effectively and enhance the diagnosis accuracy remarkably.