摘要
讨论了非线性非定常气动力的结构自适应神经网络模型建模方法,该方法具有同时进行结构辨识与参数辨识的优点;利用纵向大振幅强迫振荡风洞试验数据,验证了建模方法及所建模型的有效性,结果表明:结构自适应神经网络模型对非定常气动力有很好的逼近能力,由于采用飞行参数的时间离散数据作为输入量,模型可直接应用于飞行仿真研究。
The structure self-adapting artificial neural network method in modeling of nonlinear and unsteady aerodynamic forces is discussed in this article. This method has the virtue of doing structure and parameter identifications in the same time. The aerodynamic data from the large amplitude constrained oscillation wind tunnel experiments was used to verify the modeling method and its validity. The results show that the prediction for the unsteady aerodynamics is accurate and the model can be directly applied to the flight simulation research by taking the time discrete data as the input parameters.
出处
《飞行力学》
CSCD
北大核心
2007年第4期13-16,共4页
Flight Dynamics
基金
国防装备预研基金资助项目(513130103)
关键词
非定常气动力
气动建模
神经网络
飞行仿真
unsteady aerodynamics
aerodynamic modeling
neural network
flight simulation
