A novel PT-SMC based resilient control of cyber-physical robotic system under malicious-threats

To address the problem of predefined time (PT) guaranteed performance control for uncertain robotic cyber physical systems (CPSs) under multiple cyber-attacks, this paper proposes a sliding-mode (SM) resilient control strategy with predefined time convergence (PTC) and guaranteed performance. This strategy aims to enhance the resilience of CPSs against diverse cyber-attacks (e.g., denial-of-service (DoS) and false data injection (FDI)) and improve trajectory tracking precision. First, the impact mechanisms of common cyber-attacks on CPS control signals are analyzed, and these attacks are categorized into multiplicative and additive types with respect to the control signals. Subsequently, an initial value Error Conversion Function (ECF) is designed to map arbitrary initial tracking errors into a prescribed neighborhood, thereby addressing the limitation of traditional prescribed performance control (PPC) strategies where parameter design is dependent on initial error magnitudes. Finally, for robotic CPSs under malicious cyber-attacks, an improved extreme learning machine (ELM) is employed to approximate the time-varying lumped component comprising uncertainties and cyber-attacks. By integrating the PTC sliding mode surface (SMS) and an enhanced PPC strategy, an initial-state-independent prescribed performance SM resilient control strategy is designed for robotic CPSs, which ensures control performance regardless of initial conditions. Theoretical analysis confirms the PTC of the closed-loop system. Numerical simulation results demonstrate that, under varying attack frequencies and amplitudes, the trajectory tracking errors (TTE) of the manipulator across all initial states converge within the specified time frame in accordance with the prescribed performance, achieving a tracking precision of 0.00013 rad. These findings validate the proposed algorithm's strong robustness against cyber-attacks and its practical applicability in engineering scenarios.