基于神经网络补偿动态逆误差的导弹控制器设计  

Missile Controller Design Based on Neural Network Compensating Dynamic Inverse Error

作　　者：常怡鹏 王小平[1] 林秦颖[1] 王晓光 董昕瑜 CHANG Yipeng;WANG Xiaoping;LIN Qinying;WANG Xiaoguang;DONG Xinyu(Aeronautical Engineeing College Air Force Engineering University,Xi’an 710038,China)

机构地区：[1]空军工程大学航空工程学院,西安710038

出　　处：《弹箭与制导学报》2020年第1期150-154,159,共6页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：航空科学基金(20145196023)资助。

摘　　要：针对采用动态逆方法设计导弹姿态控制器产生逆误差的问题,文中提出了一种基于神经网络在线补偿动态逆误差的控制器设计方案。首先根据奇异值摄动理论将控制系统分为慢回路和快回路,设计了两个回路相应的姿态控制器来解决导弹的强非线性以及强耦合性的控制问题,并采用神经网络在线补偿逆误差。仿真结果表明,在气动参数摄动的情况下,加入神经网络补偿逆误差的控制器能够有效的降低姿态角的跟踪误差,提高了控制系统的鲁棒性。Aiming at the problem that the inverse error of missile attitude controller was produced by dynamic inverse method,this paper proposed a controller design scheme based on neural network online compensation dynamic inverse error.Firstly,according to the singular value perturbation theory,the control system was divided into slow loop and fast loop.Two loop corresponding attitude controllers were designed to solve the problem of strong nonlinearity and strong coupling of missiles,and neural network was used to compensate the inverse error.online.The simulation results showed that in the case of aerodynamic parameter perturbation,the controller that adds neural network to compensate the inverse error can effectively reduce the tracking error of the attitude angle and improve the robustness of the control system.

关 键 词：导弹 非线性动态逆控制 快、慢回路控制器 神经网络 在线补偿 

分 类 号：TJ765[兵器科学与技术—武器系统与运用工程] V249[航空宇航科学与技术—飞行器设计]