Event-Triggered Fixed-Time Orbital Tracking Control of Spacecraft via Fuzzy Reinforcement Learning

This article addresses the fixed-time stable relative orbital tracking control problem of spacecraft by adopting a fuzzy reinforcement learning (RL) approach with an event-triggered mechanism. An identifier-critic-actor RL framework is established to integrate performance assessment and controller design. First, identifier fuzzy logic systems (FLSs) are utilized to approximate nonlinear uncertainties in spacecraft dynamic models. Subsequently, an optimal performance index is formulated by considering both system tracking errors and control efforts. Based on the cost evaluation of critic FLSs, a fixed-time optimal controller is derived via optimizing actor FLSs. Meanwhile, the designed event-triggered mechanism is employed to determine signal triggering instants. Rigorous stability analysis demonstrates that all closed-loop system signals achieve fixed-time convergence. Finally, numerical simulation results confirm the feasibility and excellent performance of the proposed control strategy.