针对导弹直/气复合控制系统，提出了一种基于滑模控制的自适应神经网络控制方法。首先，为提高导弹稳定性，运用滑模控制对气动力部分进行设计，并运用自适应神经网络控制算法消除系统建模中的误差影响，提高导弹过载的稳定性。其次，对直接力部分采用变结构的方法进行设计，提高导弹的响应速度，同时运用模糊控制理论减少变结构控制引起的抖振问题，进一步提高导弹的稳定性。实验结果表明，该控制方法能够有效提高导弹的响应速度，增加导弹的稳定性及过载能力，验证了设计方法的有效性。

An adaptive neural network control method based on sliding mode control is proposed for the reaction-jet/aerodynamic compound control system of the missile. Firstly, to improve the stability of the missile, the aerodynamic part is designed by using sliding mode control, and the adaptive neural network control algorithm is used to eliminate the error in system modeling, so that the stability of missile overloading can be improved.Secondly, the variable structure method is adopted to design the reaction-jet part to improve the response speed of the missile.Meanwhilethe fuzzy control method is used to reduce the buffeting problem caused by the variable structure control, so that the stability of the missile can be further improved.The experimental results show that the proposed control method can effectively improve the response speed of the missile, and improve the stability and overload capacity of the missile. Therefore, the effectiveness of this method is verified.