基于一致性的多智能体动态编队方法

A Consistency-Based Approach to Dynamic Formation of Multi-Agents

摘要: 近年来，多智能体的协同控制在机器人等领域已经引起了很多学者的广泛关注，并得到了迅速的发展。本文针对无人机在飞行过程当中只有一个固定编队的问题展开研究，通过对算法进行改进从而实现无人机群的动态编队，以达到解决在多种情况下的不能实现自适应问题的目的。本文基于传统的多智能体编队方法，通过引入新变量的方法进行改进，首先引入了一个时间变量，在状态量后分别加入与时间相关的函数X(t)、Y(t)，通过改变参数对编队进行仿真，使编队中心随时间的改变而不断变化，编队形式也随之改变，从而实现动态编队。最后仿真结果表明改进的算法能够实现无人机群的动态编队控制，且改进后的动态编队与原编队相比，应用范围更加广泛，更加具有普遍性。

Abstract: In recent years, the cooperative control of multi-agent in robotics and other fields has attracted extensive attention of many scholars, and has been developed rapidly. In this paper, we study the problem that there is only one fixed formation in the flight process of UAV, and improve the algorithm to realize the dynamic formation of UAV group, in order to solve in many cases can not achieve the goal of adaptive problem. In this paper, based on the traditional multi-agent formation method, a new variable is introduced to improve the formation process. Firstly, a time variable is introduced, and then time-dependent functions X(t) and Y(t) are added after the static volume , respectively, by changing the parameters to simulate the formation, the formation center changes with the time, and the formation forms also change, so as to achieve dynamic formation. The simulation results show that the improved algorithm can realize the dynamic formation control of UAV group, and the improved dynamic formation has wider application range and more universality than the original formation.