炸弹弹道落点参数拟合算法

A Combined Algorithm for Ballistic Parameters of Bomb Trajectory

针对很宽大范围抛射条件下炸弹的标准弹道系数库,提出了炸弹弹道落点参数的一种组合拟合算法。该组合拟合算法由二元插值法与神经网络法组成。首先,采用BP神经网络对炸弹弹道落点进行离线拟合,由于抛射条件范围很宽大造成了一定的拟合误差;然后,利用二元插值实时在线修正神经网络拟合所产生的计算误差。采用此组合算法对某种型号航空炸弹的弹道落点进行实时在线运算测试,通过测试实例说明,应用此组合算法进行弹道落点在线实时运算不但实时在线特性良好、拟合精度高,而且具有运算简单、结果可靠的特性。

A combined algorithm is presented for ballistic parameters fitting of bomb trajectory with very large range of the initial release conditions.This data fitting algorithm is the combination of bivariate interpolation and neutral network algorithm.The relationship between ballistic trajectory parameters（terminal time and gliding flight distance） and bomb releasing initial conditions can be described with threshold and weight of neural network, and the errors created due to large range of initial release conditions, are corrected by bivariate interpolation.The combined algorithm was used for online calculation test of the releasing point of a certain type aerial bomb, and the result showed that the algorithm can implement online calculation of the releasing point in real time.A great deal of numerical simulation results show that the combined algorithm is simple and reliable, and has fine real-time performance and high fitting accuracy.