神经网络在飞行器航迹仿真计算中的应用

in calculation of aircraft’s flight track simulation

摘要:

在飞行器航迹计算中,与插值法存在的缺陷相比,神经网络方法具有显著的优越性。通过对插值法和BP神经网络方法原理和工作过程的详细分析,以及在此基础上针对某飞行器航迹参数的计算实例,比较了线性插值、多项式拟合与神经网络方法的误差,表明神经网络方法具有更好的计算精度。

Abstract:

Compared with the limitations of interpolation method in calculation of aircraft’s flight track, Neural Network method has some notable advantages. By detailed analysis on calculating principle and process of interpolation and Neural Network, and a calculation example based on flight parameters of a aircraft, the calculation error of linear interpolation, multinomial matching and Neural Network are compared, the results show that Neural Network method has better calculation precision.