Wang Jiaqi, Gao Jinfeng, Wu Ping
School of Mechanical Engineering and Automation, Zhejiang Sci-Tech University, Hangzhou 310018, China.
School of Mechanical Engineering and Automation, Zhejiang Sci-Tech University, Hangzhou 310018, China.
ISA Trans. 2022 Sep;128(Pt A):10-16. doi: 10.1016/j.isatra.2021.10.030. Epub 2021 Nov 8.
This paper concentrates on attack-resilient event-triggered formation control of multi-agent systems (MASs) under periodic Denial-of-Service (DoS) attacks, which involves a new technique based on complex Laplacian. Firstly, periodic DoS attacks are modeled as Bernoulli distribution where the considered plant is viewed as a stochastic system. Then, a novel attack-resilient event-triggered mechanism (ARETM) is put forward to address formation shape problems of the system. ARETM not only restrains periodic DoS attacks effectively, but also saves network resources greatly and excludes Zeno behavior naturally. Moreover, the system stability is analyzed by using a generalized Nyquist stability criterion, in which a distinctive homotopy Newton iterative algorithm is introduced. Finally, numerical simulation cases are presented to explain the effectiveness of the designed method.
本文聚焦于多智能体系统（MASs）在周期性拒绝服务（DoS）攻击下的抗攻击事件触发编队控制，其中涉及一种基于复拉普拉斯的新技术。首先，将周期性DoS攻击建模为伯努利分布，在此情况下所考虑的被控对象被视为一个随机系统。然后，提出了一种新颖的抗攻击事件触发机制（ARETM）来解决系统的编队形状问题。ARETM不仅能有效抑制周期性DoS攻击，还能极大地节省网络资源并自然地排除芝诺行为。此外，利用广义奈奎斯特稳定性判据对系统稳定性进行分析，其中引入了一种独特的同伦牛顿迭代算法。最后，给出数值仿真案例以说明所设计方法的有效性。
