Deng Chao, Yue Dong, Che Wei-Wei, Xie Xiangpeng
IEEE Trans Neural Netw Learn Syst. 2022 Jun 8;PP. doi: 10.1109/TNNLS.2022.3176392.
In this article, a learning-based resilient fault-tolerant control method is proposed for a class of uncertain nonlinear multiagent systems (MASs) to enhance the security and reliability against denial-of-service (DoS) attacks and actuator faults. With the framework of cooperative output regulation, the developed algorithm consists of designing a distributed resilient observer and a decentralized fault-tolerant controller. Specifically, by using the data-driven method, an online resilient learning algorithm is first presented to learn the unknown exosystem matrix in the presence of DoS attacks. Then, a distributed resilient observer is proposed working against DoS attacks. In addition, based on the developed observer, a decentralized adaptive fault-tolerant controller is designed to compensate for actuator faults. Moreover, the convergence of error systems is shown by using the Lyapunov stability theory. The effectiveness of our result is examined by a simulation example.
本文针对一类不确定非线性多智能体系统（MASs），提出了一种基于学习的弹性容错控制方法，以增强抵御拒绝服务（DoS）攻击和执行器故障的安全性和可靠性。在所提出的协同输出调节框架下，所设计的算法包括设计一个分布式弹性观测器和一个分散式容错控制器。具体而言，通过使用数据驱动方法，首先提出一种在线弹性学习算法，用于在存在DoS攻击的情况下学习未知外系统矩阵。然后，提出了一种能抵御DoS攻击的分布式弹性观测器。此外，基于所设计的观测器，设计了一种分散式自适应容错控制器来补偿执行器故障。而且，利用李雅普诺夫稳定性理论证明了误差系统的收敛性。通过一个仿真例子验证了我们结果的有效性。
