摘要
无人机队列在运行过程中会受到风力变化、空气湍流等多种外部干扰的影响,导致无人机的飞行状态偏离预期轨迹,影响无人机队列的协同性和稳定性。为了确保队列运行的安全性,提出一种基于强化迭代学习的分布式无人机队列控制方法。结合分布式无人机队列的通信无向图获取队列中各个无人机的运行轨迹误差,并通过迭代学习算法调整无人机队列期待姿态角,实现轨迹误差补偿。在此基础上,利用强化学习调整控制参数,通过试错和学习找到最优的控制参数,使分布式无人机队列可以应对各种不确定性和干扰,实现分布式无人机队列控制。实验结果表明,所提出的方法对分布式无人机队列的机间距离和运行轨迹的控制精度较高、收敛速度较快,可以有效控制分布式无人机队列的飞行轨迹,保证列队的平稳运行。
In the operational process,drone queue may encounter various external disturbances such as wind changes and air turbulence,leading to deviations from the expected flight trajectory of drone,and affecting the coordination and stability of the drone queue.In order to ensure the safety of the queue operation,a distributed drone queue control method based on reinforcement iterative learning is proposed.By combining the communication undirected graph of the distributed drone queue,the trajectory errors of each drone in the queue are obtained,and the expected posture angles of the drone queue are adjusted through an iterative learning algorithm to compensate for trajectory errors.Building upon this,reinforcement learning is used to adjust the control parameters,the optimal control parameters are found through trial-and-error and learning,enabling the distributed drone queue to cope with various uncertainties and interferences to achieve distributed drone queue control.Experimental results demonstrate that the proposed method exhibits high control precision in terms of inter-drone distances and operational trajectory,with quick convergence speed,thus effectively controls the flight trajectory of the distributed drone queue and ensures smooth operation of the queue.
作者
孙文峰
何晓伟
SUN Wenfeng;HE Xiaowei(Eductional Technology Center,Shanghai Jian Qiao University,Shanghai 201306,China)
出处
《微型电脑应用》
2025年第12期287-290,298,共5页
Microcomputer Applications
关键词
强化迭代学习
分布式无人机
队列控制
期待姿态角
参数整定
reinforcement iterative learning
distributed drone
queue control
expected posture angle
parameter tuning
