拦截弹道快速设计方法  

Rapid design algorithm for intercept missile trajectory

作　　者：王备[1] 周韬[1] 董长虹[1] 

机构地区：[1]北京航空航天大学宇航学院,北京100191

出　　处：《北京航空航天大学学报》2015年第2期358-363,共6页Journal of Beijing University of Aeronautics and Astronautics

摘　　要：针对传统优化算法在解决多约束条件下拦截弹道设计问题时迭代计算耗时过长的缺点,提出了一种快速拦截弹道设计方法.该方法基于人工神经网络的非线性拟合技术,先通过普通优化算法进行离线样本计算,对样本中目标位置和优化得到的飞行程序参数进行输入-输出映射拟合训练,从而实现在给定目标位置的情况下对最优飞行程序参数的快速计算,达到提高拦截弹道设计效率的目的.经过数值仿真验证,将该方法应用于拦截弹道设计,在保证良好命中精度的同时,设计所需的计算时间大大减少,从而提高了弹道设计的时效性.Traditional optimization algorithm always needs iterative calculation and consumes long time when it is used to solve the problem of intercept trajectory design under the multi-constraints condition. In view of this disadvantage,a rapid trajectory design algorithm for intercept missile was proposed. The new algorithm was based on the nonlinear fit technique of artificial neural network. First,a sample was calculated offline through a general optimization algorithm. The target position and the optimized flight procedure parameters in the sample were trained to fit the input-output mapping. Then the optimal flight procedure parameters could be calculated fast in the condition that the target position was given,and the goal of improving the trajectory design efficiency was achieved. Applying this algorithm to intercept trajectory design,while ensuring good hit accuracy,the calculation time is greatly reduced compared with the traditional algorithm,and it improves the timeliness of intercept trajectory design. The numerical simulation demonstrates the effectiveness of the algorithm proposed.

关 键 词：拦截弹 弹道 优化 设计 神经网络 

分 类 号：V412.41[航空宇航科学与技术—航空宇航推进理论与工程]