航空学报 > 2020, Vol. 41 Issue (1): 323316-323316   doi: 10.7527/S1000-6893.2019.23316

无人机桨叶损伤的在线模型估计新方法

张霄, 倪铭, 余翔, 郭雷   

  1. 北京航空航天大学 自动化科学与电气工程学院, 北京 100083
  • 收稿日期:2019-07-25 修回日期:2019-08-23 出版日期:2020-01-15 发布日期:2019-09-16
  • 通讯作者: 余翔 E-mail:10296@buaa.edu.cn
  • 基金资助:
    国家重点研发计划(2017YFB0503000);国家自然科学基金(61627810,61833013)

A novel estimation method for blade impairments of unmanned aerial vehicle

ZHANG Xiao, NI Ming, YU Xiang, GUO Lei   

  1. School of Automation Science and Electrical Engineering, Beihang University, Beijing 100083, China
  • Received:2019-07-25 Revised:2019-08-23 Online:2020-01-15 Published:2019-09-16
  • Supported by:
    National Key R&D Program of China (2017YFB050300); National Natural Science Foundation of China (61627810,61833013)

摘要: 无人机在复杂高对抗环境中极易出现桨叶结构受损故障,造成无人机控制性能退化甚至失稳,给无人机带来灾难性后果。桨叶结构损伤条件下无人机动力学模型的在线估计与重建是保障无人机控制系统稳定的重要前提。由于桨叶损伤干扰隐含于无人机动力学模型的内部,可观测性较低,典型的干扰观测器难以实现对此类干扰的估计。提出了一种新型穿透型干扰观测器,通过构造穿透函数,将模型内部的干扰映射到一个新建的平行空间,实现了对桨叶损伤的估计和故障后无人机的建模,并给出了所设计穿透型干扰观测器的稳定性条件。以四旋翼无人机为应用对象,对桨叶损伤形成的干扰进行在线估计与量化,反演出桨叶随机出现的损伤,实现了桨叶损伤后无人机动力学模型的在线精细重建,解决了无人机桨叶损伤故障下力矩输入难以直接测量的问题。半物理仿真实验验证了所提方法的有效性。

关键词: 在线建模, 模型不确定量化, 干扰估计, 桨叶损伤, 无人机

Abstract: Highly complicated environments usually lead to blade impairment of unmanned rotor aerial vehicle,causing deleterious impact on control performance and stability, even inducing catastrophic consequences. Online estimate and reestablishment of the dynamic model in the presence of blade impairment are solid foundations of ensuring the stability of the post-fault unmanned rotor aerial vehicle. Nevertheless, since the blade impairments are inside the unmanned rotor aerial vehicle, its low visibility renders the typical observer method unable to achieve estimation. This study presents a novel penetrating disturbance observer that can estimate the blade damage via the establishment of a penetrating function and the projection of the blade damage. Moreover, the convergence of estimation error is verified and the model uncertainty is quantitatively analyzed. The hardware-in-the-loop platform based on a quadrotor unmanned aerial vehicle is established to validate the effectiveness of the proposed method. The experiment results exemplify that the developed method can estimate the disturbance induced by blade impairments, overcoming the difficulty of estimating torque under the condition of blade impairment, thereby achieving the model recognition.

Key words: online modeling, quantitative analysis of model uncertainty, disturbance observer, blade impairment, unmanned aerial vehicle

中图分类号: