研究了应用于具有超机动能力的导弹的非线性神经网络控制系统 .利用神经网络可以任意逼近非线性函数的能力 ,来学习超机动导弹在大迎角状态下的高度非线性动态特性的逆动态特性 ,得到对参数变化及未建模动态具有较好鲁棒性的控制系统 .为了改善神经网络的学习能力和学习算法的稳定性 ,对学习增益做了模糊化处理 .数字仿真结果说明了该控制方案可以达到预期效果 . The nonlinear neural network control system applied to missiles with over maneuver capability is studied.The neural network can be used to approximate the nonlinear function of the missile to learn the inverse dynamics of the highly nonlinear dynamic characteristics of the hyper-maneuverable missile at high angle of attack The control system with good robustness to parameter changes and unmodeled dynamics is obtained.In order to improve the learning ability of neural network and the stability of the learning algorithm, the learning gain is fuzzified.The numerical simulation results show that the Control program can achieve the desired results.