Error-Triggered Flexible Prescribed Performance Control for Free-Flying Space Robot Stochastic Systems with Actuator Faults

This paper presents a flexible prescribed performance control (PPC) strategy for free-flying space robots (FFSRs) under system uncertainties, external disturbances, output constraints, and actuator faults. Firstly, the control strategy is formulated by developing a stochastic model for FFSRs, wherein the stochastic neural networks (SNNs) are incorporated to estimate the lumped stochastic variable standing for both system uncertainties and exterior perturbations. Then, a class of error-triggered mechanism (ETM) related to PPC is proposed. The ETM is designed to act solely on the single prescribed performance boundary (PPB) and will activate immediately just when the tracking error approaches the PPB. Furthermore, by redesigning the PPC, the proposed control strategy removes the initial feasibility condition (IFC) while simultaneously addressing settling time specifications, asymmetrical regulating, and steady-state error modification. The devised control approach guarantees that all signals in the closed-loop FFSR stochastic system remain semi-globally uniformly ultimately bounded in probability (SGUUBIP), while the tracking errors converge to any predefined accuracy in a specified time. In conclusion, simulation outcomes for FFSR demonstrate the proposed method's effectiveness and superiority.