【作者】 刁琛；

【导师】 鲜斌；

【作者基本信息】 天津大学 ， 控制理论与控制工程， 2013， 博士

【摘要】 四旋翼无人飞行器是一种典型的欠驱动系统，所谓欠驱动特性是指在系统中控制输入变量数目少于系统自由度数目，这种特性虽然可以有效降低系统的设计与制造难度，但是给系统的控制设计带来了很大的挑战。此外四旋翼无人飞行器动力学模型还具有非线性、强耦合以及静不稳定等特点，加之飞行器在飞行过程中容易受到非线性的空气阻力、阻力矩以及阵风等不确定因素的影响，所以控制设计具有很大的难度。四旋翼无人飞行器具有垂直起降、操作灵活以及精确悬停等特点，已经被广泛用于军事以及民用等领域。但是操作四旋翼无人飞行器要求操纵人员具有丰富的遥控经验，因此为保障四旋翼无人飞行器的飞行安全，提高飞行任务的效率，需要研究具有高精度和高可靠性的飞行控制策略。四旋翼无人飞行器位姿控制是利用飞行器在飞行过程中的实时位姿信号设计反馈控制器，使飞行器对预定的位置和姿态轨迹进行跟踪，并且由于飞行器机载驱动器输出功率的限制，控制输出必须保持在合理的范围以内。近年来，国内外许多科研人员对于四旋翼无人飞行器位姿控制进行了广泛研究，提出了多种控制策略，但是这些方法都存在一些不足，因此目前四旋翼无人飞行器位姿控制问题尚没有得到完全的解决。本文针对具有建模不确定性的四旋翼无人飞行器系统进行了深入研究，并设计了数种性能可靠的控制策略。针对四旋翼无人飞行器动力学模型中存在未知参数的情况，论文采用速度有界PID算法以及自适应控制算法设计了一种调节控制器，该控制器可以保证飞行器快速准确地调节至指定位置与航向。由于飞行器在飞行过程中容易受到阵风等外界扰动的影响，论文利用非线性滑模控制算法与自适应算法设计了一种自适应滑模控制器，这种飞行控制器可以实现对四旋翼无人飞行器三个方向的位置与偏航角的跟踪控制。论文采用反步法并结合非线性鲁棒算法提出了一类光滑的非线性跟踪控制算法，该算法可以有效消地除传统滑模控制器中存在的颤振问题。由于四旋翼无人飞行器自身负载能力有限，无法携带高精度传感器，论文针对无人飞行器位姿控制问题，采用单位四元数姿态描述方法，在部分飞行状态不可测量的约束条件下，设计了一种输出反馈跟踪控制器。论文利用基于Lyapunov的稳定性分析、Barbalat引理以及零扰动定理等对所提出的控制器进行稳定性分析。同时，为了验证控制器的控制效果，论文采用Matlab仿真软件对所提算法进行仿真测试，并且得到了较好的控制效果。更多还原

【Abstract】 The quadrotor unmanned aerial vehicle (UAV) is a typical underactuated system, which has less control inputs than the degrees of freedoms. Although this property can simplify the design and manufacture, it brings a great challenge to the control development. The dynamics of the quadrotor UAV also suffers from various control complexities such as: strongly coupling and nonlinearity. Besides, the quadrotor UAV is effected by the nonlinear aerodynamic drag during the flight. These dynamic uncertainties complicate the development of the controller.Due to its advantages such as vertical taking off and landing (VTOL), rapid maneuvering and precise hovering, the potential for the quadrotor UAV in civil and military applications has been well established. However, a lot of experience is required for the human pilot. High performance control laws are required for the autonomous flying control of the quadrotor UAV. Meanwhile, in consideration of the limitation of onboard actuators, the control inputs should be with some reasonable values. Recently, several literatures have proposed some new methods for the control of the quadrotor UAV, but the design of nonlinear control mechanisms for the quadrotor UAV in the presence of structural uncertainties and unknown external disturbances is still a challenging task.A further investigation on controller design for the quadrotor UAV is proposed, and several kinds of control algorithms are developed in this dissertation. In the chapter II, a new nonlinear adaptive controller is proposed for a class of quadrotor UAV, which is subjected to parametric uncertainties. The on-line parameter estimations are combined with the feedback control to develop the nonlinear adaptive regulation controller, which yields a global asymptotic regulation result for the UAV’s Cartesian position and yaw angle while keeps the closed system stable. In the chapter III, an adaptive sliding mode tracking controller is proposed for the quadrotor UAV which is subject to parametric uncertainties and external disturbances. By using Lyapunov-based stability analysis, it can be proved that the proposed control law yields a global asymptotic tracking result for the UAV’s Cartesian position and the yaw orientation. In the chapter IV, on-line parameter estimations are combined with a robust control design based on hyperbolic tangent functions to develop a new continuous and differentiable tracking controller, which yields an asymptotic tracking result for the UAV’s Cartesian position and an asymptotic regulation result for the yaw orientation. Because of the load capacity of the UAV, some high accuracy sensors may not apply to this class of aircraft. In the chapter V, an attitude and altitude output feedback tracking controller is developed, which uses only the system’s output state and yields a semi-global asymptotic tracking result. With the aid of Lyapunov-based stability analysis, Barbalat’s Lemma and vanishing perturbation theorem, the stability of the closed-loop system can be guaranteed for the proposed control strategies. Numerical simulation results are included to validate the performance of the presented control laws.更多还原