摘要
目前的航空发动机气路故障诊断主要依靠人工测量与巡检,这种方式无法将与时间相关的发动机参数变量综合考虑,诊断中主观性较强,造成过剩维修与故障漏检的问题,提出基于关联挖掘的航空发动机故障诊断方法,建立发动机的截面温度、压力、转子转速、燃油流量等一些变量的历史数据库,对数据进行特殊预处理后,防止脏数据对模型的建立带入大量的噪声,使用关联挖掘技术建立故障诊断知识库,解决了在故障诊断中因为人工主观性出现过诊断的现象,用测试数据对模型进行检验证明这种模型对故障的诊断准确率高达85%,对发动机智能故障诊断系统设计有很强的借鉴意义。
The aeroengine gas path fault diagnosis relies mainly on the manual measurement and inspection, this approach cannot be re lated to time engine parameter variables are taken into account, in the diagnosis of subjectivity is stronger, excess maintenance and fault un- detection problem, aircraft engine fault diagnosis method based on association mining is put forward, build engine section temperature, pres- sure, some variables, such as rotor speed, fuel flow history database, special after preprocessing the data, to prevent the dirty data for the establishment of the model into a lot of noise, and USES association mining technology to establish fault diagnosis knowledge base, solved because of human subjectivity in the fault diagnosis of diagnosis, with test data to test the model proved that the model of fault diagnosis accuracy is as high as 85 ~, the engine intelligent fault diagnosis system design has a strong reference value.
出处
《计算机测量与控制》
北大核心
2014年第2期379-381,389,共4页
Computer Measurement &Control
基金
河南省教育厅科学技术研究重点项目(12A520054)
关键词
关联规则
测量参数
性能参数
气路参数
Association rule
Parameter measurement
Performance parameter
Gas path parameters
