具有复杂障碍物的多智能体蜂拥控制算法

Flocking Control Algorithm of Multi-Agent System with Complex Obstacles

摘要: 本文研究具有切换拓扑网络的一般二阶多智能体系统蜂拥控制算法，当多智能体系统在遇到复杂障碍物时，引入了一种新的虚拟智能体产生一个切向排斥力，使得多智能体系统能够更光滑的通过障碍物。同时，采用动态牵制策略控制整个网络的连通性，确保多智能体系统能够达到蜂拥状态且智能体之间不发生碰撞。最后给出了一些数值模拟结果，证实了算法的有效性。

Abstract: The flocking control algorithm for general second-order multi-agent systems with switching topology networks is investigated in this paper. When the multi-agent system encounters complex obstacles, a new virtual agent is introduced to produce a tangential repulsion force, so that the multi-agent system can pass through obstacles more smoothly. At the same time, the dynamic pinning control strategy is adopted to control the connectivity of the whole network, so as to ensure that the multi-agent system can reach the flocking state and there is no collision between agents. Finally, some numerical simulation results are given to verify the effectiveness of the algorithm.