Abstract: The traditional distributed formation control method of Virtual Structure exhibited limited flexibility in transformation, was unable to avoid obstacles in time. To solve this problem, a Unmanned Aerial Vehicle (UAV) formation control algorithm based on virtual control network was proposed. For global path planning, the individual drones in the formation maintained fixed positions and flew in a diamond formation as a whole, following the virtual center point. In the local path planning, the Artificial Potential Field method was used to avoid known obstacles. For unknown static obstacles in the environment, the gravity of the anchor point and the target point were add to improve the Artificial Potential Field method, so that the UAV flew around the unknown obstacle and returned to the original path after successfully avoiding obstacle. Simulation results show that the Improved Artificial Potential Field method has shorter length compared to the A* algorithm, Rapidly-exploring Random Tree algorithm (RRT), and Dynamic Window Approach (DWA) in the path planning. In terms of runtime, it saves 81.27% compared to Rapidly-exploring Random Tree Star (RRT*), 48.3% and 46.97% compared to DWA and RRT*, which proves the effectiveness of the Improved Artificial Potential Field method.

Key words: Unmanned Aerial Vehicle &#40, UAV&#41, formation control, path planning

摘要： 针对传统虚拟结构的无人机分布式编队控制方法队形变换存在灵活性不足，遇到障碍物躲避不及时等问题，提出一种基于虚拟控制网络的无人机编队控制算法。全局路径规划时，编队中各无人机在队形中占据固定位置，跟随虚拟领航者以整体菱形队形飞行。局部路径规划中，采用人工势场法进行已知障碍物避障，对于环境中的未知静态障碍物，通过引入锚点引力以及目标点引力改进人工势场法，使无人机编队围绕未知障碍物飞行，并在避障成功后返回原路径。仿真实验结果表明，改进的人工势场法的路径长度均小于A*算法、快速探索随机树算法(RRT)、动态窗口算法(DWA)；在运行时间上，比快速探索随机树*(RRT*)算法时间节省81.27%，比动态窗口算法和快速探索随机树算法分别节省48.33%和46.97%，证明了所提算法的有效性。

关键词: 无人机编队控制, 路径规划, 虚拟结构法, 人工势场法