基于遗传神经网络的反跑道子弹药落点散布研究

摘要

为研究布撒器反跑道子弹药时序落点散布,并实现对子弹药时序落点的精确预估,分析了子弹药的运动模型以及飞行时序,影响子弹药落点的因素主要有投放初始条件、机构误差和风干扰.采用了神经网络与遗传算法相结合的方法,建立了落点散布的预估模型和仿真程序.仿真结果表明,该方法能够有效地描述反跑道子弹药时序落点散布,同时可以进行精确预估计算,可为布撒器的总体设计提供重要技术手段.%To study time-sequence impact point dispersion of anti-runway s ubmunitions（ARSM）, and to precisely estimate impact point,the motion model and flight time-sequence of ARSM were analyzed. The main factors affecting impact point are initial condition, mechanism error and wind disturbance. The method of combining neural networks and genetic algorithms was designed. The estimating models of impact point dispersion of submunition were established,and the simulation procedure was designed. The simulation results show that the method can effectively describe time-sequence impact point dispersion of ARSM, and accurate pre-estirnation can be carried out by this method. The method offers important technical means for the overall design of dispenser.