Riaz Saleem, Li Bingqiang, Qi Rong, Zhang Chenda
School of Automation, Northwestern Polytechnical University, Xi'an, 710072, China.
Sci Rep. 2024 Mar 28;14(1):7361. doi: 10.1038/s41598-024-57775-8.
Malicious attacks are often inevitable in cyber-physical systems (CPS). Accuracy in Cyber physical system for position tracking of servos is the major concern now a days. In high precision industrial automation, it is very hard to achieve accuracy in tracking especially under malicious cyber-attacks, control saturations, parametric perturbations and external disturbances. In this paper, we have designed a novel predefined time (PDT) convergence sliding mode adaptive controller (PTCSMAC) for such kind of cyber physical control system. Main key feature of our control is to cope these challenges that are posed by CPS systems such as parameter perturbation, control saturation, and cyber-attacks and the whole system then upgrade to a third-order system to facilitate adaptive control law. Then, we present an adaptive controller based on the novel PDT convergent sliding mode surface (SMS) combined with a modified weight updated Extreme Learning Machine (ELM) which is used to approximate the uncertain part of the system. Another significant advantage of our proposed control approach is that it does not require detailed model information, guaranteeing robust performance even when the system model is uncertain. Additionally, our proposed PTCSMAC controller is nonsingular regardless of initial conditions, and is capable of eradicating the possibility of singularity problems, which are frequently a concern in numerous CPS control systems. Finally, we have verified our designed PTCSMAC control law through rigorous simulations on CPS seeker servo positioning system and compared the robustness and performance of different existing techniques.
在网络物理系统（CPS）中，恶意攻击往往不可避免。如今，网络物理系统中伺服器位置跟踪的准确性是主要关注点。在高精度工业自动化中，尤其是在恶意网络攻击、控制饱和、参数摄动和外部干扰的情况下，很难实现跟踪的准确性。在本文中，我们为这类网络物理控制系统设计了一种新颖的预定义时间（PDT）收敛滑模自适应控制器（PTCSMAC）。我们控制的主要关键特性是应对CPS系统带来的这些挑战，如参数摄动、控制饱和和网络攻击，然后将整个系统升级为三阶系统以促进自适应控制律。然后，我们提出一种基于新颖的PDT收敛滑模面（SMS）并结合改进的权重更新极限学习机（ELM）的自适应控制器，该极限学习机用于逼近系统的不确定部分。我们提出的控制方法的另一个显著优点是它不需要详细的模型信息，即使在系统模型不确定时也能保证鲁棒性能。此外，我们提出的PTCSMAC控制器无论初始条件如何都是非奇异的，并且能够消除奇异性问题的可能性，而奇异性问题在众多CPS控制系统中经常是一个关注点。最后，我们通过对CPS导引头伺服定位系统进行严格仿真验证了我们设计的PTCSMAC控制律，并比较了不同现有技术的鲁棒性和性能。
