在全局环境未知且存在静态障碍物的情况下,将群体扩展为具有有限记忆能力的智能群体,研究了该群体的群集运动及其避障行为。模型利用智能个体记忆单元的信息实时产生动态切换的拓扑网络结构;在人工势能结合速度一致策略的基础上,采用滚动窗口方法引入虚拟领航者,利用极限环方法解决群体避障问题,实现了快速且安全的群集运动控制。仿真结果验证了该模型的有效性。 When the global environment is unknown and there are static obstacles, the group is expanded into a group of intelligent individuals with limited memory. The cluster movement and its obstacle avoidance behavior are studied. The model uses the information of intelligent individual memory unit to generate dynamically switched topological network structure in real time. On the basis of the consistent strategy of artificial potential energy, a rolling window method is used to introduce the virtual leader, and the limit cycle method is used to solve the obstacle avoidance problem in population. And safe cluster motion control. The simulation results verify the effectiveness of the model.