论文部分内容阅读
针对变后掠翼近空间飞行器(near space vehicle,NSV)在大包络、多任务模式飞行运动过程中具有非线性、快时变、强耦合和不确定的特性,提出了基于径向基神经网络(radialbasis function neural network,RBFNN)的鲁棒自适应跟踪控制策略。首先,利用RBFNN在线逼近NSV飞行过程中外部干扰。其次,应用backstepping设计光滑的反馈控制器。其中,采用微分器避免backstepping设计中出现微分膨胀问题,利用鲁棒项减少RBFNN估计误差对系统的影响。然后,通过公共Lyapunov函数证明所提出的控制器可以保证在任意飞行模态中NSV的输出跟踪误差均可以收敛到任意小的有界集内。最后,仿真结果表明该飞控系统具有良好的控制性能。
Aiming at the nonlinear, fast time-varying, strongly coupled and uncertain characteristics of near space vehicle (NSV) in large enveloping and multi-mission flight, Robust adaptive tracking control strategy for radialbasis function neural network (RBFNN). First, RBFNN is used to approximate the external disturbance during the flight of NSV. Secondly, we design a smooth feedback controller using backstepping. Among them, the use of differentiator to avoid backstepping design differential expansion problems, the use of robust reduction RBFNN estimation error on the system. Then, it is proved by the common Lyapunov function that the proposed controller can ensure that the output tracking error of NSV can converge to any small bounded set in any flight mode. Finally, the simulation results show that the flight control system has good control performance.