面向网络攻击下弹性均值趋同的智能体群组编队

Group Agent Formation for Resilient Average Consensus under Attacks

中文关键词  多智能体系统  鲁棒性  弹性均值趋同  编队

英文关键词  Multi-Agent Systems  Network Robustness  Resilient Average Consensus  Formation

中文摘要

      本文研究通信范围有限的智能体群组编队问题, 探索网络攻击下多智能体系统弹性均值趋同控制策略. 现 有的工作表明, 多智能体系统可通过维持一个所谓r–鲁棒的通信网络, 实现分布式弹性趋同控制器的设计. 然而, 传 统的方法中只有当单个智能体的通信范围足够大时, 才能使网络满足r–鲁棒这一特定条件. 本文利用智能体可移动 的特性放宽了对通信范围的要求, 通过小组化和模块化的设计思想以及相应的编队策略, 让智能体沿预设轨道做周 期性运动, 从而达成具有r–鲁棒的通信网络, 并基于该通信网络提出了一种分布式弹性均值趋同控制方法. 此外, 分 析给出了单模块和多模块情况下多智能体网络达成r–鲁棒的所需条件. 最后在仿真实例中, 实现了网络攻击下系统 状态的均值趋同, 实验结果验证了群组编队和均值趋同控制方法的有效性.

英文摘要

      In this paper, we study the problem of group agent formation with limited communication range, and explore the resilient average consensus control strategy for multi-agent systems under attacks. Existing work shows that multi-agent systems can realize the design of distributed resilient consensus controller by maintaining a so-called r-robustness communication network. However, in the traditional methods, only when the communication range of a single agent is large enough, can the network meet the specific condition of r-robustness. In this paper, we relax the requirements on communication range by leveraging the agent motion, and through the grouping and modular design and the corresponding formation strategy, the agents are allowed to move periodically along the preset track, so as to achieve an r-robustness network, and based on this network, a distributed resilient average consensus control law is proposed. In the simulation example, the average consensus of the system under attacks was realized, and the experimental results verified the effectiveness of the group agent formation and the proposed average consensus control method.