基于自适应动态逆的自主飞艇速度控制系统设计  

Velocity Control System of Autonomous Airship Based on Adaptive Dynamic Inversion

作　　者：郭建国[1] 

机构地区：[1]西北工业大学精确制导与控制研究所,西安710072

出　　处：《宇航学报》2008年第5期1505-1508,1514,共5页Journal of Astronautics

摘　　要：针对无人自主飞艇上升或下降的速度控制问题,提出了一种基于神经网络自适应动态逆控制的飞行速度控制设计方法。基于自主飞艇浮力控制系统的配置状况,建立了飞艇飞行速度的数学模型。引入速度控制系统的参考模型,并利用神经网络学习功能,获得非线性控制系统的动态逆;通过引入速度误差的非线性补偿项,自适应补偿整个控制系统由于动态特性变化引起的误差,克服包括参数摄动与外界扰动在内的各类不确定性因素的影响,实现飞艇对期望速度的稳定跟踪。最后仿真结果表明,所提出的控制方案对无人自主飞艇的飞行速度具有良好的控制效果,从而验证了控制系统设计的有效性。A kind of design method of velocity control system is proposed based on dynamic inversion and neural network, according to the problem of velocity control for unmanned autonomous airship. Duo to the scheme of buoyancy control system in the airship, the flight kinematics model was established based on forces acted on the airship. After a reference model was introduced, the dynamic inversion of nonlinear control system was obtained through a training neural network. Applied the nonlinear error information, the errors caused by the inversion and the dynamic characteristic of control system were compensated adaptively. Finally an illustrative example is given to show that the proposed control scheme can adapt well to large-scale variation of model parameter perturbation and wind turbulence, and the valid of velocity control system is verified.

关 键 词：速度控制 动态逆 神经网络 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]