摘要
针对冲突信息条件下的故障诊断难题,研究了基于近似表示空间的故障诊断方法;采用冲突诊断信息系统描述冲突故障信息,并将冲突诊断信息系统转化为近似表示空间,进而对其进行属性约简;设计了基于包含度的针对完备信息对象的故障诊断方法,采用向量补齐方法设计了针对不完备信息对象的近似表示空间故障诊断方法;最后,将该方法应用于航空发动机故障诊断,验证了该方法可有效处理冲突、不完备信息,显著提高了故障诊断准确率。
A fault diagnostic approach is proposed to deal with inconsistent information. Firstly, inconsistent diagnostics information system is introduced to express the inconsistent information. And the inconsistent diagnostics information system is translated into consistent approximate denote space. Then knowledge reduction is carried out. Secondly, diagnostic method based on inclusion degree is designed for complete-information objects. And vectors are complemented for incomplete--information objects. At last, this approach is validated with aero-engine data. Result showed that the designed approach can deal with incomplete and inconsistent information effectively, and enhance the accuracy of diagnosis remarkably.
出处
《计算机测量与控制》
CSCD
北大核心
2012年第6期1583-1586,共4页
Computer Measurement &Control
基金
空军工程大学研究生创新基金(Dx2010107)资助课题
关键词
故障诊断
属性约简
冲突信息系统
diagnosis
attribute reduction
inconsistent information system