本文为美国密苏里科技大学（作者：DAVID JOHN NODLAND）的硕士论文，共106页。

本文研究直升机无人机的最优控制问题。直升机无人机可广泛用于军事和民用领域。由于这些直升机是欠驱动非线性机械系统，高性能控制器的设计是一个挑战。本文提出了一种基于神经网络的直升机无人机轨迹跟踪控制器的状态反馈和输出反馈优化设计方法。状态和输出反馈控制系统基于后退式方法，采用运动学和动力学控制器，而输出反馈方法除了这些控制器外还使用了一个观测器。基于在线逼近器的动态控制器在连续时间内学习Hamilton-Jacobi-Bellman（HJB）方程，计算相应的最优控制输入，使HJB方程在时间上向前最小化。利用单一的神经网络进行代价函数逼近，实现最优跟踪。利用Lyapunov分析证明了系统的整体闭环稳定性。仿真结果验证了所提出的轨迹跟踪控制设计的有效性。还包括用于确认理论方法的硬件描述，以及与所使用的算法和特定于硬件实现的方法有关的材料讨论。此外，还特别关注实施过程中的挑战以及在这一领域进行进一步研究的机会。本文以两篇论文的形式提出了上次研究观点。

This thesis addresses optimal control of ahelicopter unmanned aerial vehicle (UAV). Helicopter UAVs may be widely usedfor both military and civilian operations. Because these helicopters areunderactuated nonlinear mechanical systems, high performance controller designfor them presents a challenge. This thesis presents an optimal controllerdesign via both state and output feedback for trajectory tracking of ahelicopter UAV using a neural network (NN). The state and output-feedbackcontrol system utilizes the back stepping methodology, employing kinematic anddynamic controllers while the output feedback approach uses an observer inaddition to these controllers. The online approximator-based dynamic controllerlearns the Hamilton Jacobi-Bellman (HJB) equation in continuous time andcalculates the corresponding optimal control input to minimize the HJB equationforward-in-time. Optimal tracking is accomplished with a single NN utilized forcost function approximation. The overall closed-loop system stability isdemonstrated using Lyapunov analysis. Simulation results are provided todemonstrate the effectiveness of the proposed control design for trajectorytracking. A description of the hardware for confirming the theoreticalapproach, and a discussion of material pertaining to the algorithms used andmethods employed specific to the hardware implementation is also included.Additional attention is devoted to challenges in implementation as well as toopportunities for further research in this field. This thesis is presented inthe form of two papers.

1. 引言

2. 基于神经网络的无人机最优控制及其硬件实现

更多精彩文章请关注公众号：