基于BP神经网络的无动力弹攻击区拟合分析

Attack Zone Fitting of No-Power Missile Based on BP Neural Network

准确地预测导弹攻击区是提高导弹命中精度的有效方法。采用BP神经网络算法对某无动力弹的攻击点进行拟合,所以在不同的初始条件下可得出不同的可攻击区。无动力弹攻击点与攻击条件可以通过神经网络的阈值和权值来表现。通过实例说明了应用BP神经网络进行无动力弹攻击区拟合,不但算法可行性好、拟合精度高,而且运算简单。

Accurately predicting attack zone（AZ） of missile is an effective approach for improving target-aimed precision. Back propagation neural network algorithm which was applied to fit attack points of no-power missile（APNPM） was employed in the paper, therefore, different AZ can be obtained in different initial conditions. The relation between APNPM and attack conditions can be displayed with threshold and weight of BP neural network. The results of simulation show that the BP method not only has good feasibility and high precision, but also operation simple. While in the whole process of which C＋＋ and Matlab were used.