一类具有群体LEADER的多智能体系统的聚集行为

针对目前关于群体聚集的稳定性研究中仅考虑同类智能体的问题，提出了一类由lea der和follower 两种个体组成的多智能体系统聚集模型.在该模型中，只有leader智能体具有关于环境和目标点的信息，并且当个体间距离较远时互相吸引，距离较近时互相排斥.当表示环境信息的势能场函数梯度有界时，证明了系统的聚集稳定性并确定了表示群体大小的一个上界，从而保证了所有leader和follower个体最终聚集在群体中心周围并一起沿着该函数的负梯度方向运动.仿真结果验证了方法有效性.

Aggregation of Multi-Agent systems with group leaders

A new model including agents of leaders and followers is proposed,in which the leaders have the knowledge about the environment and the goal,but the followers do not.When any two of the agents are far away,they will attract each other.Otherwise,they repell.Under the assumption that the gradient of the environment represented by a potential function is bounded,it is shown that the aggregation of the multi-agent systems is stable and further,and a super limit about the size of the group is obtained.Thus,the leaders and followers can be guaranteed to aggregate around the center of the group's position and to move together along the negative gradient of the environment.The obtained results are illustrated by numerical examples.