摘要
介绍了两种炸弹弹道落点拟合神经网络模型,对各模型的优缺点以及处理结果进行了分析。提出了采用广义回归神经网络来处理炸弹弹道落点拟合问题,弹道的落点参数、初始投放条件与风场可通过神经网络的阈值和权值来表现。仿真结果表明,应用广义回归神经网络进行弹道落点拟合,具有算法可行性好、拟合精度高、速度快等优点,而且运算简单;该方法在实战中有很高的参考价值和工程实用价值。
This paper introduced two kinds of neural network algorithm to fit the bomb trajectory falling points. The merits and processing result of two models were compared and analysed. The method of generalized regression neural network（GRNN） was brought out to deal with the problem of bomb trajectory falling point fitting. The relation of trajectory falling point parameters, initial release conditions and wind fields can be displayed with threshold and weight of GRNN. The results of simulation show that the GRNN method not only has good feasibility and high precision, but also operate simply and fast. The conclusion possesses reference value and practical engineering value in real applications.
出处
《飞行力学》
CSCD
北大核心
2009年第4期46-49,共4页
Flight Dynamics
