航空学报 > 2023, Vol. 44 Issue (S1): 727645-727645   doi: 10.7527/S1000-6893.2022.27645

不确定强耦合下四旋翼姿态鲁棒自适应控制

刘晨阳1, 吴大伟1(), 郭一泽1, 吕欣赛1, 周佳妮1, 邵书义2   

  1. 1.河海大学 能源与电气学院,南京  211100
    2.南京航空航天大学 自动化学院,南京  211106
  • 收稿日期:2022-06-17 修回日期:2022-07-28 接受日期:2022-08-26 出版日期:2023-06-25 发布日期:2022-09-13
  • 通讯作者: 吴大伟 E-mail:wudaweiwkl@126.com
  • 基金资助:
    国家自然科学基金(62103135);中央高校基本科研业务费专项资金(B210202068)

Robust adaptive attitude control of quadrotor with uncertain strong coupling

Chenyang LIU1, Dawei WU1(), Yize GUO1, Xinsai LV1, Jiani ZHOU1, Shuyi SHAO2   

  1. 1.School of Energy and Electrical Engineering,Hohai University,Nanjing  211100,China
    2.College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing  211106,China
  • Received:2022-06-17 Revised:2022-07-28 Accepted:2022-08-26 Online:2023-06-25 Published:2022-09-13
  • Contact: Dawei WU E-mail:wudaweiwkl@126.com
  • Supported by:
    National Natural Science Foundation of China(62103135);Fundamental Research Funds for the Central Universities(B210202068)

摘要:

针对多源不确定强耦合下的四旋翼无人机姿态控制问题,首次设计了一种多层逼近自适应神经网络动态面控制算法。区别于以往的加性耦合不确定研究,考虑无人机飞行控制中的乘性耦合多源不确定估计与补偿问题。首先,构建不确定四旋翼无人机姿态动力学模型,并基于神经网络与傅里叶展开实现乘性耦合不确定的巧妙转换;其次,将自适应技术与反步法相结合,设计多层逼近自适应控制律;同时将动态面技术用于解决反步法中虚拟控制律求导问题。完整理论分析与仿真实验表明了所述控制策略的有效性。

关键词: 四旋翼无人机, 姿态控制, 神经网络, 动态面控制, 自适应控制

Abstract:

A dynamic surface control algorithm based on multi-layer adaptive neural network is proposed for the attitude control problem of the quadrotor UAV with multi-source strong coupling uncertainties for the first time. Different from the previous research on additive coupling uncertainties, the problem of multi-source multiplicative uncertainties estimation and compensation in UAV flight control is considered. First of all, a four-rotor UAV attitude dynamics model with multiplicative strong coupling uncertainties is constructed, and the uncertainties are ingeniously converted based on neural network and Fourier expansion. Secondly, a multi-layer approximation adaptive control law is designed based on the combination of adaptive technology and backstepping method. At the same time, dynamic surface technology is used to solve the derivation problem of virtual control law in the backstepping method. The effectiveness of the proposed control strategy is proved through complete theoretical analysis and simulation experiments.

Key words: quadrotor UAV, attitude control, neural network, dynamic surface control, adaptive control

中图分类号: