Cyber-physical systems (CPSs) seamlessly integrate communication, computing, and control, thus exhibiting tight coupling of their cyber space with the physical world and human intervention. Forming the basis of future smart services, they play an important role in the era of Industry 4.0. However, CPSs also suffer from increasing cyber attacks due to their connections to the Internet. This article investigates resilient control for a class of CPSs subject to actuator attacks, which intentionally manipulate control commands from controllers to actuators. In our study, the supertwisting sliding-mode algorithm is adopted to construct a finite-time converging extended state observer (ESO) for estimating the state and uncertainty of the system in the presence of actuator attacks. Then, for the attacked system, a finite-time converging resilient controller is designed based on the proposed ESO. It integrates global fast terminal sliding-mode and prescribed performance control. Finally, an industrial CPS, permanent magnet synchronous motor control system, is investigated to demonstrate the effectiveness of the composite resilient control strategy presented in this article.

中文翻译：

---

遭受执行器攻击的信息物理系统的复合有限时间弹性控制


网络物理系统（CPS）无缝集成通信、计算和控制，从而表现出网络空间与物理世界和人类干预的紧密耦合。它们构成了未来智能服务的基础，在工业 4.0 时代发挥着重要作用。然而，由于与互联网的连接，CPS 也遭受了越来越多的网络攻击。本文研究了一类遭受执行器攻击的 CPS 的弹性控制，这些攻击有意操纵从控制器到执行器的控制命令。在我们的研究中，采用超扭曲滑模算法构建有限时间收敛扩展状态观察器（ESO），用于估计系统在存在执行器攻击的情况下的状态和不确定性。然后，针对受攻击系统，基于所提出的ESO设计了有限时间收敛弹性控制器。它集成了全局快速终端滑模和规定性能控制。最后，对工业 CPS（永磁同步电机控制系统）进行了研究，以证明本文提出的复合弹性控制策略的有效性。