摘要
论述了基于伪控制补偿解决自适应控制中作动器饱和问题的方法。基本控制律采用非线性动态逆方法设计,神经网络用于对逆误差进行重构。伪控制补偿消除作动器和自适应单元之间的交互影响。通过在超机动飞行控制的应用仿真表明,该控制方案弥补了动态逆要求精确数学模型的缺点,消除了作动器饱和对自适应单元的影响,提高了整个控制系统的鲁棒性。
Pseudo-Control Compensation is investigated in this paper, which has been used to resolve actuator saturation in adaptive control. The base control law is designed in dynamic inversion and neural networks are used to reconstruct inversion error. Pseudo-Control Compensation cancel the interaction between the actuator and the adaptive factor. Simulations of super-maneuvering flight control show that limitation of needing accurate mathematical model in dynamic inversion method can be released and actuator affects on adaptive factor is removed in this control structure, and that the robustness of the whole system is greatly improved.
出处
《系统仿真学报》
CAS
CSCD
2003年第5期727-730,共4页
Journal of System Simulation
关键词
神经网络
自适应控制
飞行控制
动态逆
neural network
adaptive control
flight control
dynamic inversion