摘要
针对非线性控制提出一套较理想的解决方案。在动态逆控制的基础上,利用神经网络的在线学习能力建立自适应控制结构,结合滑模控制的较强的鲁棒性和瞬态性,两次互补性地补偿误差,从而提高动态逆控制的鲁棒性,解决建模精度不高的问题。仿真结果表明这种方法对高性能无人机飞行控制设计研究具有一定理论指导意义和应用价值。
A new method for nonlinear flight control systems is presented . The method is based on dynamic inversion control theory, and an adaptive control structure is built using neural network, error is compensated twice using sliding mode control which exhibits excellent robustness and transient characteristic performance. The results indicate that the method are valuable for the control system design and research in advanced UAV.
出处
《航空兵器》
2006年第4期3-6,共4页
Aero Weaponry
关键词
神经网络
自适应控制
滑模控制
非线性控制
动态逆
neural network
adaptive control
sliding mode control
nonlinear control
dynamic inversion
