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。
The flocking with obstacle avoidance in global unknown environment was studied based on intelligent agents that have limited memories. The model produced a dynamic switch networks with information in the memory space. Based on the strategy of combining artificial potential with velocity consensus, the swarm was navigated to goal by virtual leader which was produced with rolling window method, and obstacle avoidance was solved by limit-cycle method .The swarm gets to the goal rapidly and safely in the form of flocking at last. The results of simulation show that the algorithm is valid.