BPN 在涡扇发动机气路故障诊断中的应用

本文应用误差反传网络（ＢＰＮ）算法，研究发动机气路部件故障的定量诊断，并针对某型军用涡扇发动机的典型故障，给出了诊断结果，分析了测量系统的随机误差、发动机的不同工作状态、不同调节规律对的诊断结果影响，结果表明ＢＰＮ可成功地应用于发动机故障的定量诊断。

APPLICATION OF BPN TO QUANTITATIVE DIAGNOSIS OF GAS PATH COMPONENT FAULTS IN TURBOFAN

The possibility of using BPN to quantitative diagnosis of turbofan engine gas path component faults is investigated.By organizing the fault samples quantitatively,the construction of BPN is built and applied to the fault diagnosis of a low bypass ratio turbofan engine.The diagnostic results show that BPN can not only isolate the gas path component faults,but also identify the severity of component performance deterioration.It is proved that the BPN is adaptive to random errors in measure system and different engine operation regimes for a control plan.But the construction of BPN must be rebuilt under the condition of different engine control plans.Because it has a little ability of self-organization,it is diffcult for BPN to isolate the engine faults quantitatively when several faults occur at the same time.