航空学报 > 2026, Vol. 47 Issue (17): 333200-333200   doi: 10.7527/S1000-6893.2026.33200

面向狭窄通道穿越的无人机编队队形高速安全变换

边疆, 聂天常(), 李宇华, 张振华, 郭超勐   

  1. 北方自动控制技术研究所,太原 030006
  • 收稿日期:2025-12-05 修回日期:2026-01-05 接受日期:2026-03-03 出版日期:2026-03-24 发布日期:2026-03-23
  • 通讯作者: 聂天常 E-mail:nie_tianchang@163.com
  • 基金资助:
    国家自然科学基金(U24A20263)

High-speed safe formation transformation of UAVs formation through narrow corridors

Jiang BIAN, Tianchang NIE(), Yuhua LI, Zhenhua ZHANG, Chaomeng GUO   

  1. North Automatic Control Technology Institute,Taiyuan 030006,China
  • Received:2025-12-05 Revised:2026-01-05 Accepted:2026-03-03 Online:2026-03-24 Published:2026-03-23
  • Contact: Tianchang NIE E-mail:nie_tianchang@163.com
  • Supported by:
    National Natural Science Foundation of China(U24A20263)

摘要:

针对旋翼无人机(UAV)编队飞行过程中高速穿越狭窄通道并尽可能保持相近编队构型的问题,提出一种多阶段UAV编队队形安全变换方法。首先,将UAV编队中心看成虚拟领导者,根据UAV的运动学约束,通过混合整数线性规划规划出一条平滑、高速、距离障碍物最远的虚拟领导者轨迹以及编队整体的安全飞行空间;然后,根据编队安全飞行空间和虚拟领导者轨迹求解一个无外部和内部碰撞,且相较初始构型改变较少的编队构型;之后,基于生成的编队构型和虚拟领导者轨迹,分布式计算每个UAV的期望轨迹,并提出一种结合非线性优化与协同时间重分配的分布式算法,对各UAV轨迹进行优化,使其在满足运动学约束的前提下,实现高速飞行与编队保持。仿真结果表明:所提方法能使UAV编队在满足运动学约束的情况下,高速、安全穿越狭窄通道,同时尽可能保持相近的编队构型。

关键词: 无人机编队, 编队变换, 混合整数线性规划, 非线性优化, 遗传学算法

Abstract:

A multi-stage safe formation transformation method for Unmanned Aerial Vehicles (UAVs) formation is proposed to address the problem that rotary-wing UAVs formation passes through narrow corridors at high speed while maintaining a similar formation configuration as much as possible. Firstly, the center of the UAVs formation is regarded as a virtual leader. According to the kinematic constraints of UAVs, a mixed-integer linear programming method is used to plan a smooth and high-speed virtual leader trajectory with the maximum distance from obstacles and the safe flight space of the entire formation. Then, based on the safe flight space of the formation and the virtual leader trajectory, a formation configuration without external and internal collisions and with minimum changes compared with the initial configuration is obtained. Subsequently, based on the generated formation configuration and the virtual leader trajectory, the desired trajectories of each UAV are calculated in a distributed manner. A distributed algorithm combining nonlinear optimization and cooperative time redistribution is proposed to optimize the trajectories of each UAV, so that high-speed flight and formation maintenance can be achieved while satisfying the kinematic constraints. Simulation results show that the proposed method enables the UAVs formation to pass through narrow corridors safely and at high speed while satisfying the kinematic constraints, and maintains a formation configuration close to the initial configuration as much as possible.

Key words: UAVs formation, formation transformation, mixed-integer linear programming, nonlinear optimization, genetic algorithm

中图分类号: