航空发动机的健康状态是保证飞行安全的重要因素,能在早期发现发动机隐藏的故障,通过预测发动机性能参数的变化趋势来掌握发动机性能衰退情况,实现对发动机健康状态的准确判断,具有重要的意义。针对航空发动机性能参数的数据特点,提出将动态神经网络中的NARX(非线性自回归)神经网络模型运用到性能参数的预测中,并用航空发动机排气温度(EGT)的趋势预测对该方法进行了验证。验证结果表明,这种方法在性能参数预测的精度上优于BP神经网络的预测结果。%Aeroengine has been known as the heart of the aircraft ,and its performance is important in ensu‐ring flight safety .Therefore ,it is significant for aircraft safety since it is able to reflect the decline of engine performance through predicting the change trend of engine performance parameters .In view of the data characteristics of aeroengine performance parameters ,the NARX neural network model is proposed and used for the prediction of performance parameters .This article employs the model to predict the aeroengine exhausted gas temperature (EGT) .According to the results ,the NARX neural network model has a higher precision than the model based on the BP neural network in parameters change trend and data accuracy .
展开▼