超立方体在大规模的多智能机器人编队方式中的设计

Hypercube in Large-Scale Way in the Design of Intelligent Robot Team Formation

在多智能体机器人群组中,机器人的数量越多,能完成任务的复杂程度也就越高,所需的时间也就越少。但随着机器人数量的增加,系统的通信开销必然以爆发式的方式增长。一旦有机器出现故障,就有可能会导致任务失败。如何在任务复杂度和系统的通信开销这两个矛盾的变量中,达到我们所需要的最好结果。我们采用超立方体的结构化设计理念,又通过蚁群子立方体算法对超立方体进行二次划分,使多智能体机器人群体在扩张性、容错性、事务复杂度等各个方面都取得了较好的结果。

In multi-agent robot group,the number more of robots,can complete the task of the higher the complexity,the less time required for it.But with the increasing number of robots,the communication system is necessarily the way to explosive growth.Once the machine fails,it might cause the task to fail.How the task complexity and system communication variables of these two conflicts,to achieve the best results we need it.We adopt hypercube structured design concept,and then through ant colony algorithm of sub-cube for hypercube secondary division; multi-agent robot groups has better results in all aspects of the expansion,fault tolerance and transaction complexity.