基于神经网络的小型无人直升机非线性鲁棒控制设计  

Nonlinear robust control for a small unmanned helicopter based on neural network

作　　者：鲜斌 张浩楠 XIAN Bin, ZHANG Hao-nan(School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, Chin)

机构地区：[1]天津大学电气自动化与信息工程学院,天津300072

出　　处：《控制与决策》2018年第4期627-632,共6页Control and Decision

基　　金：国家自然科学基金项目(60804004;90916004);天津市应用基础与前沿技术研究计划重点项目(14JCZDJC31900);天津市科技支撑计划重点项目(15ZCZDGX00810);天津市科技计划项目(14RCHZGX00862)

摘　　要：针对小型无人直升机的姿态控制问题,为补偿系统参数不确定性和外界扰动的影响,设计一种连续的非线性鲁棒控制器.首先,利用神经网络在线估计系统不确定性,采用基于误差符号函数积分的鲁棒控制算法抑制外界扰动,同时补偿神经网络估计误差;然后,利用基于Lyapunov函数的分析方法,证明所设计控制器的闭环稳定性,确保无人直升机姿态误差的半全局渐近收敛;最后,在无人直升机飞行实验平台上,进行无人机抗风扰控制实验.实验结果表明,所提出的控制方法具有良好的控制效果,对系统不确定性和外界扰动具有良好的鲁棒性.This paper presents an asymptotic tracking controller for a small-size unmanned helicopter using neural network（NN） and a continuous nonlinear robust control strategy called robust integral of the signum of the error（RISE）.The NN in the control law approaches the uncertain parts of the helicopter＇s dynamics using online network weight tuning, while the NN approximation errors and the external disturbances are compensated by using the RISE approach.Semi-global asymptotic stability of the error signals and the boundedness of the closed-loop system signals are ensured via Lyapunov based stability analysis. Finally, real-time experiments are performed on a helicopter attitude control test-bed.The experimental results show that the proposed controller achieves a good performance and the closed loop systems obtain good robustness with respect to system uncertainties and external disturbances.

关 键 词：无人直升机 神经网络 鲁棒控制 抗扰动 渐近稳定 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]