研究了航空发动机气路故障诊断的测量参数选择的问题。基于发动机故障诊断的主因子模型，结合预测平方残差和(prediction error sum of squares，简称PRESS)准则，建立了一种选择风扇流量、风扇出口压力、压气机出口压力、压气机出口温度、涡轮后温度和燃油流量这6个测量参数进行单轴涡扇发动机故障诊断的方法。大量的实例表明，合理运用该方法选择测量参数的种类和数量，既能取得比较满意的诊断效果，又能利用发动机控制系统已有的测量参数，降低诊断成本，具有一定的工程应用价值。
Research on selection of measurement parameters of engine fault diagnosis based on primary factor model
The selection of measurement parameters of engine air path fault diagnosis was researched in this paper based on the primary factor model and prediction error sum of squares (PRESS) principle. Through plenty of computing results, a method was established for fault diagnosis of one spool turbofan engine by selecting six measurement parameters, namely, corrected mass flow of fan, outlet total pressure of fan, outlet total temperature of compressor, outlet total pressure of compressor, outlet total temperature of low-pressure turbine and fuel mass low. This can also obtain satisfactory diagnosis results while the results can also be applied to other forms of aero engines. It is indicated that the reasonable type and quantity selection of measurement parameters not only guarantees the diagnosis, but also cuts down the cost. The method is of some engineering value.
fault diagnosisprimary factor modelconjugate gradient methodPRESS(prediction error sum of squares)principle
NETL
NSTL
万方数据
维普
