Predefined-Time Actor-Critic Learning for Constrained Spacecraft Attitude Consensus

초록

This article focuses on the problem of predefined-time suboptimal attitude consensus control for multiple spacecraft with time-varying state constraints. Under the backstepping framework, a predefined-time optimized controller is developed through the actor-critic architecture, where the actor executes control actions and the critic evaluates their effectiveness. The proposed approach derives the updating laws from the negative gradient of a positive function generated by the partial derivative of the Hamilton-Jacobi-Bellman (HJB) equation, thereby avoiding the complexity of traditional Bellman residual-based methods and relaxing the persistence of excitation condition. Based on the barrier Lyapunov function, it is proved that the proposed strategies can steer the spacecrafts to attitude consensus in a predefined time, and the states of multiple spacecraft do not violate the prescribed constraints. The simulation results verify the effectiveness of the proposed approach in a predefined time.

키워드