Resilient Output Containment of Heterogeneous Cooperative and Adversarial Multigroup Systems
Resilient Output Containment of Heterogeneous Cooperative and Adversarial Multigroup Systems
复制标题
DOI：
10.1109/tac.2019.2947620
复制
发表时间：
2020-07
影响因子：
6.8
通讯作者：
Shan Zuo;F. Lewis;A. Davoudi
中科院分区：
文献类型：
--
作者：
Shan Zuo;F. Lewis;A. Davoudi
本文提出了一种新的合作对抗多群体系统的概念，该系统由合作的领导者和追随者以及对手组成。例如，多个军用车辆在复杂的动态网络环境中协同作战，其中存在未知的敌人和隐藏的恶意攻击者。在分布式环境中缺乏全局态势感知，使得自动驾驶汽车容易受到网络攻击和渗透。每个代理都不知道其邻居的动机，并且可以从队友和对手接收信息/数据。网络化多群系统需要安全的、弹性的控制协议来防止攻击者的攻击在网络中传播，从而影响系统的性能甚至整体稳定性。针对传感器故障和攻击者的攻击，提出了一种分布式弹性控制结构，保证了闭环动态系统的一致最终有界性。数值模拟结果表明了所提出的结果的有效性。
This note introduces a new concept of cooperative and adversarial multigroup system, which consists of cooperative leaders and followers, as well as adversaries. For example, cooperation of multiple military vehicles operate in complex dynamic networked environments with unknown enemies and hidden malicious attackers. The lack of global situational awareness in distributed settings makes autonomous vehicles prone to cyberattacks and infiltration. Each agent is unaware of the motives of its neighbors and may receive information/data from both the teammates and the adversaries. Secure and resilient control protocols are essential for the networked multigroup systems to prevent the adversaries’ attacks from propagating across the network, which may influence the system performance and even overall stability. To counter sensor faults and attacks from the adversaries, a distributed resilient control architecture is proposed, which guarantees uniform ultimate boundedness of the closed-loop dynamical system. Numerical simulations illustrate the effectiveness of the proposed results.