摘要
研究了发动机状态参数异常对发动机整体性能的影响,利用多元统计分析方法建立了状态变量对发动机性能影响的概率模型,用于发动机性能监控过程中的故障分离。根据模型能够计算出发动机性能出现异常时各个状态变量导致此时刻性能异常的概率,据此初步判断哪些状态变量导致发动机性能出现异常,并根据概率的大小对相应部件进行检查,以此来寻找故障源。在某型涡扇发动机性能监控中的应用表明,通过计算异常样本点上各监测参数对样本点异常的影响概率,并结合发动机故障机理分析,能够有效地确定故障源。
The influence of state parameter abnormity on aeroengine performance is studied. Based on multivariate statistical analysis a probability model describing state variable how to influence aeroengine performance is built to isolate fault in performance monitoring of aeroengine. Based on this influence probability model the probability of each state variable making performance abnormaity when abnormity accrues in aeroengine is computed. According to this probability the faulted variable which causes performance abnormity can be estimated and the relevant components are examined in order to find the potential fault of engine. The practical applications in monitoring certain type of turbo-fan engine show that fault origin can effectively be found out by calculating the probability of each monitored parameter making the sample abnormal, combined with aeroengine fault mechanism analysis.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2007年第3期483-487,共5页
Systems Engineering and Electronics
关键词
航空发动机
性能监控与故障诊断
故障分离
多元统计分析
概率模型
aeroengine
performance monitoring and fault diagnosis
fault isolation
multivariate statistical analysis
probability model