无人机技术的不断发展带动了社会生产、日常生活、军事农业等方面飞速的发展，正因如此，无人机智能自主飞行越来越成为当下研究的重要方向及热点。本文针对无人机智能自主飞行中相关的关键问题进行了研究。主要包括：无人机航迹规划方法研究及分析、无人直升机自转下滑最优控制问题研究以及无人机飞行任务效能仿真评估。在无人机航迹规划部分，本文针对不同任务的航迹规划需求，改进和提出了不同的航迹规划方法，包括：改进人工势场法、改进A*算法、新型曲线规划算法等，同时结合粒子群优化算法提出了节点粒子群路径规划方法，并与课题组实际参与的高原地区运输飞行任务当中进行的手工航迹规划结果做了对比分析，其结果表明了该方法的有效性。除此之外，还以毫米波雷达为传感器，进行了避障仿真飞行试验，成功实现了智能自主避障飞行。在直升机的自转下滑最优控制问题部分，论文提出了S-PSO算法，并优化了自转下滑问题的数学模型，针对S-58样例直升机进行了算例对比，结果证明该算法有相对较优的精确性和计算效率。在任务效能仿真评估部分论文进行了航迹规划任务的仿真，通过前面所述的航迹规划方法优化出合理航迹，并建立了相关的效能评估方法进行了任务效能评估。论文主要工作意义在于：一方面改进了现有的航迹规划方法，并完成了避障飞行试验，同时针对场地飞行任务提出了新型的曲线规划算法和改进人工势场法，而针对无人机高原运输飞行任务提出了节点粒子群算法，相比手工路径规划，航迹最高飞行高度降低200m，航迹总长度缩短了约2km，同时也提升了无人机飞行任务路径规划效率。另一方面在直升机自转下滑问题中，采用最优控制思想，融合梯度下降与随机优化的思路，提出了新的优化算法，提升了自转轨迹优化效率，能够给出最优自转飞行操纵，同时能够计算绘制直升机的低速回避区曲线，能够提升直升机自转飞行的成功率。
The continuous development of UAV technology has led to the rapid development of social production, daily life and military agriculture, etc. Because of that, intelligent and autonomous flight of UAVs has increasingly become an important direction and hotspot of the current research.The paper studies the key issues related to intelligent autonomous flight of UAVs. Mainly contents include: research and analysis of UAV trajectory planning methods, research on optimal control of unmanned helicopter autorotation, and simulation of mission effectiveness of UAV’s flight. In the UAV trajectory planning section, the paper improves and proposes different trajectory planning methods for different mission trajectory planning needs, including: improved artificial potential field method, improved A * algorithm, new curve planning algorithm, etc. Combining with particle swarm optimization algorithm, a node particle swarm path planning method is proposed and compared with the manual trajectory planning results of the plateau transportation missions actually participated by the research team. The results show the effectiveness of the method. In addition, the millimeter-wave radar is used as a sensor to carry out obstacle-avoidance simulation flight test, and successfully achieving intelligent autonomous obstacle-avoidance flight. In the part of optimal control of helicopter's autorotation, the paper proposes the S-PSO algorithm and optimizes the mathematical model of the problem of autorotation, and compares the calculation examples for the S-58 sample helicopter. The result proves that the algorithm is relatively superior accuracy and computational efficiency. In the part of the mission effectiveness simulation evaluation, the paper simulates the trajectory planning task, optimizes the reasonable trajectory through the aforementioned trajectory planning method, and establishes a related effectiveness evaluation method to perform the mission effectiveness evaluation.The main significance of the paper is: on one hand, it improves the existing trajectory planning method and completes the obstacle avoidance flight test. At the same time, it proposes a new curve planning algorithm and improved artificial potential field method for the field flight mission, while targeting the UAV The plateau transportation mission has proposed a node particle swarm algorithm. Compared with manual path planning, the maximum flight height of the flight path is reduced by 200m, and the total length of the flight path is shortened by about 2km. At the same time, the path planning efficiency of the UAV flight mission is also improved. On the other hand, in the helicopter autorotation problem, using the optimal control idea, combining the idea of ?gradient descent and stochastic optimization, a new optimization algorithm is proposed, which improves the optimization efficiency of the rotation trajectory, and this algorithm can give the optimal autorotation flight control, can draw the helicopter's low-speed avoidance zone curve, and can improve the success rate of the helicopter's autorotation flight.
