首页> 外文OA文献 >Dynamic modelling and flight control methodologies for vertical take-off and landing unmanned aerial vehicles
【2h】

Dynamic modelling and flight control methodologies for vertical take-off and landing unmanned aerial vehicles

机译:垂直起降无人机的动态建模和飞行控制方法

摘要

This thesis makes contributions to three important areas related to vertical take-off and landing (VTOL) unmanned aerial vehicles (UAVs). Firstly, a two-stage grey-box system identification approach is proposed to identify the hover dynamics of a helicopter. One of the main impediments to a proper grey-box identification is the choice of initial values of the unknown parameters, which affects the values to which the parameters converge during optimisation. The first stage of the proposed approach utilises specialised filters to find high-quality initial values in the partitioned hover model. The second stage starts with the obtained initial values and identifies the complete hover model using frequency-domain estimation methods. The approach is validated using real flight data.Secondly, a comprehensive dynamic model for a novel ducted fan vehicle is developed from first principles. The vehicle is powered by three ducted fans, two of which can be vectored and it is designed to be able to translate with minimal attitude variation, which cannot be achieved by conventional helicopters, quadrotors and single ducted fans. The resulting nonlinear model includes all the dominant elements, namely the chassis dynamics, the aerodynamic features of the ducted fans, the actuator dynamics and gyroscopic effects and is used to obtain the hover and cruise models for controller design.Thirdly, centralised robust controllers are proposed for VTOL UAVs in low-speed flight. Due to the naturally coupled dynamics of VTOL UAVs, a centralised robust stabilisation control scheme is proposed, i.e., no partitioning of the vehicle model is needed in the controller design and the robustness requirements of different subsystems are satisfied simultaneously. By using the structured H∞ synthesis, the centralised control scheme is further merged with decentralised controllers with prescribed structure and order. Two controllers with such configurations are designed separately to control the heading and position of the ducted fan vehicle against internal and external disturbances and they also enable forward flight with minimal pitch variation, which is further utilised in the ``turn-and-cruise'' way-point navigation flight. Simulation results are provided to illustrate the validity of the controllers.
机译:本文为与垂直起降(VTOL)无人机(UAV)有关的三个重要领域做出了贡献。首先,提出了一种两阶段灰箱系统识别方法来识别直升机的悬停动力学。正确识别灰箱的主要障碍之一是未知参数初始值的选择,这会影响参数在优化过程中收敛到的值。所提出方法的第一阶段利用专用滤波器在分区悬停模型中找到高质量的初始值。第二阶段从获得的初始值开始,并使用频域估计方法识别完整的悬停模型。第二,从第一性原理出发,开发了一种新型管道风机的动力学模型。该车辆由三个风管风扇提供动力,其中两个可以进行矢量控制,并且设计为能够以最小的姿态变化平移,这是常规直升机,四旋翼飞机和单风管风扇无法实现的。由此产生的非线性模型包括所有主要因素,即底盘动力学,管道风扇的空气动力学特性,执行器动力学和陀螺效应,并用于获得用于控制器设计的悬停和巡航模型。第三,提出了集中鲁棒控制器适用于低速飞行中的VTOL无人机。由于VTOL无人机的自然耦合动力学特性,提出了集中式鲁棒稳定控制方案,即在控制器设计中不需要划分车辆模型,并且同时满足了不同子系统的鲁棒性要求。通过使用结构化的H∞综合,集中控制方案进一步与具有规定结构和顺序的分散控制器合并。具有这种配置的两个控制器是分开设计的,以控制风车的航向和位置不受内部和外部干扰的影响,它们还能够以最小的俯仰变化实现向前飞行,这在``转弯巡航''中得到了进一步的利用航点导航飞行。提供仿真结果以说明控制器的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号