神经网络在直升机空空导弹攻击中的应用研究

应用神经网络方法研究了直升机发射空空导弹的一些问题。文中首先介绍了直升机空空导弹攻击的火控系统原理；其次，对复杂且不稳定的直升机飞行运动模型设计了模糊神经网络控制器；然后，用BP网络实时计算了直升机发射空空导弹的空中攻击包线；最后，设计了综合火／飞控制系统，并对所设计的整个攻击系统进行了仿真，通过仿真证明满足系统设计要求。在神经网络设计中，采用了变尺度优化算法，提高了算法的速度和精度，对空战的实时应用提供了重要的参数价值。

Application Study of Neural Network in Attacking of Air-Air Missiles Launched by Helicopter

Some problems about air air missiles launched by helicopter are studied using neural networks. Firstly, the principle of fire control system about air air missiles launched by helicopter is presented. Secondly, since the armed helicopter model is complex and has less stability, a fuzzy neural network controller (FNNC) is designed. Thirdly, the launch envelop's real time calculation is studied using BP neural networks. Finally, the whole process of designed system is simulated. Result shows that the design requirements are met. In the design of neural network, optimizing algorithm of DFP is used and reference value for real time application in aerial combat is provided.