四旋翼无人机轨迹跟踪与编队控制方法研究
文献类型：学位论文
| 作者 | 杜晗1,2
|
| 学位类别 | 工学硕士 |
| 答辩日期 | 2017-05 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 北京 |
| 导师 | 易建强 |
| 关键词 | 四旋翼无人机 轨迹跟踪 编队控制 反馈线性化 自抗扰控制 |
| 中文摘要 | 四旋翼无人机以其体积小、机械结构简单、控制代价小、可垂直起降、能自主悬停等优势，在军事和民用领域都得到了广泛的应用。对于其自主轨迹跟踪控制系统的研究，是保证其安全飞行、适应多变应用场景、顺利完成任务的关键。进一步地，多四旋翼无人机系统有着单四旋翼难以企及的优势，通过编队协同适应更多应用场景、覆盖范围更广、工作效率更高，有着深远的研究意义。本文在深入研究四旋翼模型特性的基础上，设计了鲁棒的轨迹跟踪控制系统和编队控制方案，并进行了室外飞行实验验证。本文主要工作包括： （1）建立了四旋翼无人机的精确动力学模型方程，包括旋翼动力学模型、机体线运动、角运动与地效应建模。特别地，在研究中对陀螺效应与地效应等复杂动力学特性进行了详细的定性与定量分析，并说明了其对于四旋翼无人机动力学特性的显著性影响。 （2）针对四旋翼模型耦合程度高、欠驱动、模型不确定性高及易受外部扰动的特点，设计了一种包含水平位置、姿态角、内环等三个控制回路的轨迹跟踪控制系统。该系统采用反馈线性化方法设计位置、姿态角两个回路的控制器，采用增量式动态逆方法设计内环控制器，以应对陀螺效应给控制系统设计带来的困难。此外，基于自抗扰理论设计了安排过渡过程与扩张状态观测器，以提高系统的鲁棒性。最后，利用分离原则证明了反馈线性化控制器与扩张状态观测器组成的控制-观测系统的稳定性。 （3）针对多四旋翼编队协议与四旋翼底层非线性动力学模型难以结合，以及编队控制过程中不可避免的轨迹跟踪误差、编队扰动的问题，提出了一种基于虚拟结构法的编队控制系统设计方案。该方案中，首先设计了虚拟领航者、领航者信息拓扑与多四旋翼通信拓扑；其次，基于反馈线性化方法将各个位置编队通道解耦为二阶模型，基于一致性理论设计出编队协议；再次，通过设计受限指令滤波反步控制器跟踪作为编队协议虚拟控制量的升力、姿态控制指令；最后，结合自抗扰控制进行编队扰动观测器设计，估计并消除外界扰动，提升编队控制系统的鲁棒性。 （4）在仿真实验验证四旋翼无人机轨迹跟踪与编队控制方法有效性的基础上，通过基于自主搭建的多四旋翼飞行实验平台的一系列室外飞行实验进行了进一步验证。轨迹跟踪飞行实验中，无人机实现了圆形和“8”字形的精确轨迹跟踪。多机编队实验飞行方面，双机编队变换、三机战术模拟等多种构型编队实验也取得了较好的结果。 |
| 英文摘要 | Quadrotor Unmanned Aerial Vehicles (QUAV) have been widely utilized for both military and civilian purposes owing to its small size, simple structure, little control cost, vertical takeoff and land (VTOL) , autonomous hovering and other superiorities. The research on its trajectory tracking control guarantees QUAV to fly safely, adapt to variable situations and achieve their goals. What’s more, Multiple Quadrotor System (MQS) has great advantage over single one on its adaptations to more difficult situations, broader cover range, and efficiency relying on cooperation. Based on the comprehensive analysis of the dynamic characteristics of QUAV, a robust trajectory tracking control system and an integrated formation control scheme are designed and tested by outdoor experiments. The main contributions are addressed as follows. (1) Accurate dynamic equations of QUAV are derived, including propeller dynamic model, body line motion model, angular motion model and the ground effect. Particularly, gyro effect and ground effect are analyzed in detail from qualitative and quantitative points of view to show their significant influence on the dynamic model. (2) To cope with the problem that the QUAV model is an under-actuated, highly coupled nonlinear system which is sensitive to disturbances, a three-loop trajectory tracking control system is designed. The system is composed of horizontal position subsystem, attitude subsystem, and inner subsystem. The feedback linearization (FL) theory is utilized in the design of controllers for both horizontal position and attitude subsystems. A controller based on incremental nonlinear dynamic inversion is proposed to cope with the special form when gyro effect is considered in the inner subsystem. Arranged transient process (ATP) and extended State observers (ESO) are designed for the system to improve its robustness against disturbances. Finally, the stability of the control-observe cascade system composed of FL and ESO is proven based on separation principal. (3) To cope with the problem that formation protocol is difficult to be combined with bottom dynamic nonlinear model, suffering from unavoidable trajectory tracking errors and formation disturbances, an integrated formation control scheme is proposed relying on virtual structure theory. Firstly, a virtual leader, a leader information broadcast topology, and a multi-QUAV communication topology are established. Secondly, each channel of the formation control system is decoupled into a second-order system, which is controlled by coherence protocol. Thirdly, the desired attitude angles derived from the formation protocol are tracked by a back-stepping controller with constrained command filters. Finally, ESO is incorporated in the formation control scheme to estimate and eliminate time-varying formation disturbances, thus promote the robustness. (4) With the foundation of a series of successful demonstration simulations on effectiveness and robustness of trajectory tracking and formation control, a platform for multi-QUAV outdoor experiments is established and demonstrates the effectiveness of the control scheme further. The “8” type and round trajectory are tracked accurately in experiments. Furthermore, the formation of double-QUAV and triple-QUAV systems also performs well. |
| 学科主题 | 运动稳定性与控制 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/14629]  |
| 专题 | 毕业生_硕士学位论文 |
| 作者单位 | 1.中国科学院大学 2.中国科学院自动化研究所 |
| 推荐引用方式 GB/T 7714 | 杜晗. 四旋翼无人机轨迹跟踪与编队控制方法研究[D]. 北京. 中国科学院研究生院. 2017. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。