基于交换网络的多智能体系统可控性研究综述
A Review on Controllability of Multi-Agent Systems using Switched Network
摘要:
可控性是指多智能体系统可以使用指定的规则从一种状态转向另一种状态的情况。因此,人们相信通过明确的进步来实现给定的条件。拓扑的动态性和环境的确定性是决定多智能体系统可控性的两个基本标准。一个强大的多智能体系统的拓扑结构会定期变化,改变智能体之间的连接,从而改变它们的合作努力。本调查的重点是在一个切换网络的MAS的可控性与领导者,遵循最近邻合作规则。leader/pioneer是一个单独的agent,作为输出来控制其他agent/members。由于在非确定性情况下活动的结果是未知的,Agent必须在观察其先前行动的后果后选择新的活动,这导致了时间延迟,限制了Agent的主动性。在文献中,可控性通常是以集中的方式实现的,某个领导者教育支持者如何实现特定的目标。可控性有不同的应用,包括管理飞机,车辆和机器人。
英文摘要:
Controllability refers to a situation in which a Multi-agent System may be steered from one state to another using specified rules. As a result, there is belief in achieving a given condition by explicit advances. The level of dynamism in the topology and the level of determinism in the environment are two fundamental criteria that determine multi-agent system controllability. The topology of a powerful multi-agent system changes on a regular basis, altering the connections between agents and hence their cooperative effort. This survey focuses on the controllability of MAS in a switching network with a leader that follows the closest neighbour collaboration rule. The leader/pioneer is a single agent that functions as an output to control other agents/members. Because the results of activities are unknown under non-deterministic situations, agents must choose new activities after observing the aftereffects of their prior actions, which causes time delay and limits agent proactivity. Controllability is often achieved in a concentrated manner in the literature, where a certain leader educates supporters how to achieve a specific goal. Controllability has different applications which incorporates managing airplane, vehicle, and robots.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。