Abstract: To overcome the problem which robot swarms are difficult in achieving and maintaining stabilized formations, a formation control algorithm of robot swarms based on community and multilevel topology is presented by referencing the concepts of community division and pining control. Based on the hierarchical clustering algorithm, the scale of the robot group is divided into several communities and a fixed pining robot is selected in each community. A multilevel topology structure is designed as the standard shape of formation in each community. A formation controller is proposed to finish formation and obstacle avoiding. A shape regulation force and an orientation control force are used to solve the local minimum problem and the instability of the formation direction in communities, respectively. Simulation result verifies the validity of this algorithm.

Key words: robot swarms, community division, pining control, multilevel topology, formation control

摘要： 为解决机器人群体难以形成与保持稳定队形的问题，通过引入社团划分和牵制控制思想，提出了一种基于社团和多级拓扑的机器人群体编队控制算法。基于层次聚类算法，将机器人群体分成若干社团并找出每个社团中的固定牵制机器人；设计一种多级拓扑使其成为各社团中机器人的基准队形；设计了一种机器人群体编队及避障控制器，控制器中形状调节力和方向控制力分别解决了社团中可能出现的局部极小值和队形的结构方向不确定问题，实现了群体队形的形成与保持。仿真实验证明了该算法的有效性。

关键词: 机器人群体, 社团划分, 牵制控制, 多级拓扑, 编队控制