飞机飞行振动预计技术研究

介绍了飞机飞行振动的主要来源和特征,分析了传统预计方法的不足,针对飞机飞行振动与高度、马赫数和攻角等飞行参数呈非线性关系的特点,提出了基于BP神经网络的飞行振动预计新技术,建立了预计模型,通过8组建模样本训练网络,确定了预计模型中的各个参数值。预计结果和实测飞机飞行振动信号比较分析表明,该方法预计精度高,验证了BP神经网络预计飞机飞行振动的可行性。

Study of Aircraft Flight Vibration Prediction Technology

The main source and character of aircraft flight vibration were introduced.Deficiencies of traditional prediction methods were analyzed.A new vibration prediction technology of aircraft flight based on back propagation neural network was proposed according to the non-linear relation between aircraft flight vibration and flight parameter such as Mach number,altitude,and attack angle.The prediction model was established and the value of each parameter in the model was determined by eight item training of the network.Result of comparison between aircraft flight vibration field measurement value and predicted value showed that the prediction method has good precision.So,aircraft flight vibration by using BP neural network is validated to be feasible.