摘要
针对高超声速飞行器部分状态发生不可测量问题,提出基于观测器的系统控制器设计。将所设计的被观测系统在某平衡点处线性化,用Luenberger观测器理论设计观测器增益,得到渐近收敛于系统真实状态的状态观测值。采用状态观测器与控制器分开设计的方法,应用反演控制技术,引入指令滤波器,使得伪控制量的导数可以轻易获得,简化了设计。针对系统中存在高频未建模动态,采用RBF神经网络对其进行逼近,以增强系统的跟踪性能。利用Lyapunove稳定性理论保证闭环系统有界且跟踪误差收敛于原点附近的小邻域内。通过在六自由度模型上的仿真,表明了所提方法的可行性与有效性。
Under the condition of some states of hypersonic aircraft unmeasured, a controller based on ob- server is developed. The observation system is designed through linearization in a balance position to get the states of asymptotically approach to the real states with the gains of the nonlinear observer obtained by using Luenberger observer theory. The nonlinear observer and controller are designed respectively. By introducing a command filter to modify back-stepping control, the derivatives of virtual control are easy to obtain, which greatly facilitates control design. In order to enhance the system tracking performance, The RBF neural net- work is introduced to compensate the influence from the uncertainty. The closed-loop system is guaranteed to be bounded and the tracking errors are also guaranteed to be converged exponentially to a small neighbor- hood at zero through Lyapunov stability theorem. Finally, the results are verified by simulation for a 6DOF hypersonic aircraft. The simulation shows the feasibility and validity of the proposed approach.
出处
《航天控制》
CSCD
北大核心
2012年第5期3-8,共6页
Aerospace Control
