基于边界值定量计算的神经网络自适应飞行控制  

Design of an Effective Neural Network(NN) Adaptive Flight Control Based on Unmatched Uncertain System

作　　者：李雪松[1,2] 李颖晖[2] 阎晓鹏[3] 方群[1] 李霞[2] 

机构地区：[1]西北工业大学航天学院,陕西西安710072 [2]空军工程大学工程学院,陕西西安710038 [3]北京航空工程技术研究中心,北京100076

出　　处：《西北工业大学学报》2011年第3期486-490,共5页Journal of Northwestern Polytechnical University

摘　　要：针对一般多输入多输出不确定系统,设计了一种基于自适应神经网络误差观测器控制方法。通过对一般多输入多输出系统进行匹配与非匹配两种不确定性分析,采用自适应神经网络对不确定因素导致的误差进行逼近补偿,采用Lyapunov定理证明了误差收敛有界,并计算了误差的边界值。最后将设计的控制律应用到不确定飞行器的指令跟踪控制中。仿真结果证明了该方法的有效性。Aim. The introduction of the full paper reviews a number of papers in the open literature and then, in its third and final paragraph, outlines our design, which we believes is better than can be designed with the method proposed in the 2008 Ref. 6 ; of course, starting from Ref. 6, our method makes some improvements. Sections 1, 2, and 3 explain our design ; their core consists of： （ 1 ） We describe the multiple-input multiple-output system and propose the error observer of the system; （2） We analyze its matched and unmatched uncertainty, propose the adaptive NN control method, and compensate for its uncertainty with NN; （3） We prove the convergence and the boundedness of the system error with Lyapunov theorem, and calculate the bounded set of the error. Section 4 gives a flight numerical example; the simulation results, given in Figs. 1 （a） through 1 （c） ,and their analysis show preliminarily that our proposed method can make the aircraft with uncertainties follow the commands well.

关 键 词：飞行器 神经网络 控制 设计 误差观测器 非匹配不确定 

分 类 号：V249[航空宇航科学与技术—飞行器设计]