基于回归神经网络的非线性舰炮弹道偏差预测

THE FORECAST OF NONLINEAR BALLISTIC TRAJECTORY SYSTEM BIAS BASED ON RECURSION NETWORKS

提出一种基于回归神经网络及弹道机理模型对非线性弹道系统进行辩识的方法 ,通过观测上升段弹丸飞行轨迹 ,采用混合模型预测弹道的误差规律 ,最后估计出弹着点 ,对系统采用 MATLAB进行仿真 ,结果表明 ,该方法有效可行 .

This paper proposes a method based on recursion networks for the identification of non linear ballistic trajectory processes. Via the detected climbing trajectory, using hybrid model to forecast the error rule of ballistic trajectory, and estmates the point of fall in the end. Final it is Simulated by MATLAB.