Robust Prescribed Performance Control Design Against Network-Induced Data Losses

This paper addresses the challenges posed by data losses in networked control systems (NCSs) within the framework of prescribed performance control (PPC). Data losses occurring in both the system-to-controller and controller-to-system communication channels can significantly degrade tracking performance and potentially induce internal instability, rendering conventional PPC strategies inadequate. To overcome this limitation, we propose a novel control architecture that enhances the robustness of PPC against data losses. The proposed scheme utilizes the nominal PPC controller when the tracking error remains well within the prescribed performance bounds and smoothly transitions to a robustifying mode as the error approaches critical regions of the performance envelope. Under this dual-mode operation, it is proven that the tracking error re-enters and remains strictly within the user- defined performance bounds in finite time. The approach ensures closed-loop stability and maintains tracking performance to a large extent, despite the presence of data losses. Notably, the method requires no data-loss detection or compensation mechanisms and preserves the low-complexity nature of PPC, avoiding reliance on specialized sensing or actuation hardware. Simulation results are provided to demonstrate and validate the theoretical developments.