基于试飞数据的发动机瞬态过程预测模型

Engine Transient Process Prediction Model Based on Test Flight Data

针对发动机瞬态过程的趋势监控需求,基于某型发动机飞行试验数据,采用人工神经网络建模方法,建立了该型发动机瞬态过程预测模型,并利用飞行试验数据对模型进行了验证。结果表明:模型预测结果与飞行试验数据相对误差最大值小于5%,预测结果精度较高,建模方法正确,可以用于该发动机飞行试验趋势监控,也可为其它型号发动机建模提供参考;神经网络内部神经元数量对于模型预测精度影响较大,内部神经元数量应根据最简单神经网络结构及最高模型预测精度的准则进行确定。

According to the trend monitoring demand of engine transient process,based on the flight test data of a certain type of engine,the artificial neural network modeling method was used to establish the transient process prediction model of the engine,and the model was verified by flight test data.The results show that the relative error between the model prediction result and the flight test data is less than 5%,the prediction result is accurate,and the modeling method is correct.It can be used for the trend monitoring in the engine flight test,and can also provide reference for other types of engine modeling.The number of neurons in the neural network has a great influence on the prediction accuracy of the model,the number of internal neurons should be determined according to the simplest neural network structure and the highest model prediction accuracy.