引用本文:侯捷,陈谋,刘楠.基于径向基函数神经网络与扩张状态观测器的无人直升机控制[J].控制理论与应用,2021,38(9):1361~1371.[点击复制]
HOU Jie,CHEN Mou,LIU Nan.Unmanned helicopter control based on radial basis function neural network and extended state observer[J].Control Theory and Technology,2021,38(9):1361~1371.[点击复制]
基于径向基函数神经网络与扩张状态观测器的无人直升机控制
Unmanned helicopter control based on radial basis function neural network and extended state observer
摘要点击 2846  全文点击 1038  投稿时间：2020-09-17  修订日期：2021-03-02
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2021.00624
  2021,38(9):1361-1371
中文关键词  无人直升机  飞行控制  动态面控制  神经网络  扩张状态观测器
英文关键词  unmanned helicopter  flight control  dynamic surface control  neural networks  extended state observer
基金项目  国家自然科学基金项目(61803207)、江苏省重点研发计划（社会发展）项目(BE2020704)和江苏省“333高层次人才培养工程”科研项目(BRA2019051)资助
作者单位E-mail
侯捷 南京航空航天大学 Hjsmail95@163.com 
陈谋* 南京航空航天大学 chenmou@nuaa.edu.cn 
刘楠 南京航空航天大学  
中文摘要
      针对具有系统不确定和外部干扰的无人直升机飞行控制问题, 提出了一种基于神经网络和扩张状态观测 器的控制方法. 利用神经网络逼近系统的不确定性, 引入扩张状态观测器对神经网络的逼近误差和系统外部干扰 进行估计. 基于神经网络和扩张状态观测器的输出, 对无人直升机的主旋翼挥舞角、姿态角速率、姿态角、速度与位 置系统分别进行了控制器设计, 以增强系统鲁棒性和抗干扰能力. 同时, 引入动态面控制方法以避免对虚拟信号进 行直接求导, 并通过李雅普诺夫方法分析了闭环控制系统的稳定性. 最后使用无人直升机数据进行仿真验证, 结果 表明设计的控制律能使无人直升机有效跟踪控制指令, 具有良好的稳定性与鲁棒性.
英文摘要
      In this paper, a flight control combining neural network and extended state observer is proposed for the unmanned helicopter with system uncertainty and external disturbance. Neural network (NN) is used to approximate the uncertainty of all sub-systems. The extended state observer (ESO) is introduced to estimate the approximation error of the NN and the external disturbance of the system. Based on the output of NN and ESO, the controller of main blade flapping motion, attitude rate, attitude angle, velocity and position systems of unmanned helicopter are designed respectively to enhance the robustness and anti-disturbance ability of the system. Meanwhile, the dynamic surface control method is introduced to avoid the direct derivation of virtual signals. The stability of the control law is analyzed via Lyapunov’s stability theorems. Finally, using the data of an unmanned helicopter to simulate the whole system, the results show that the designed control law can make the unmanned helicopter effectively track the control com-mands, and show good stability and robustness.