摘要
基于改进的BP神经网络方法建立了飞机在大攻角机动状态下的非定常气动力模型。分别利用X-31A绕其俯仰轴作振荡运动和F-18作大振幅谐波振荡运动的试验数据验证了模型的有效性。结果表明,建立的神经网络模型具备优良的逼近非线性非定常气动力的性能,可以有效地预测飞机大攻角非定常气动力;建立的模型具备构建多变量的大攻角非定常气动力模型的能力,且为飞机多自由度耦合的非线性非定常气动特性分析奠定了基础。
A model of unsteady aerodynamics for an aircraft at high angle of attack was set up with improved BP neural network. The validation of the model was verified with experimental data from pitching oscillation of X - 31A and large amplitude harmonic oscillation of F - 18. The results show that the model has a good capability of approaching to nonlinear unsteady aerodynamics and can effectively predict unsteady aerodynamics for the aircraft at high angle of attack. The model can set up the mathematical model of high - angle - of - attack unsteady aerodynamics, which provides a basis for studying nonlinear unsteady aerodynamic characteristics of the aircraft with more coupled freedom degrees.
出处
《计算机仿真》
北大核心
2017年第2期106-109,155,共5页
Computer Simulation
关键词
神经网络
飞机
大攻角
非定常
气动力
Neural network
Aircraft
High angle of attack
Unsteady
Aerodynamics