Actor-critic-based predefined-time control for spacecraft attitude formation system with guaranteeing prescribed performance on SO(3)

Abstract

In this paper, an attitude formation control problem of multi-spacecraft systems is investigated on the special orthogonal group (SO(3)) under the undirected tree communication topology. A novel appointed-time performance function is developed to characterize the user-assigned transient and steady state performance. To online compensate the system uncertainties, an actor-critic neural network architecture is utilized in the control design process. The critic neural network intends to approximate the long-term integral cost function, which can evaluate the consensus performance of the formation system. Based on the exported reinforcement signal, the actor neural network is introduced to generate the feedforward compensation term to cope with the system uncertainties. By utilizing the barrier Lyapunov function, a model-free decentralized control scheme is constructed with an adaptive robustifying term to enhance the disturbance rejection ability, in which the control input for each spacecraft is produced by using only the local relative information from its neighbors. Moreover, an auxiliary system is devised to cope with the input saturation constraint. By using the Lyapunov stability theory, the stability and convergence of the closed-loop system are analyzed. Finally numerical simulations are conducted to illustrate the feasibility and effectiveness of the proposed adaptive actor-critic learning-based control scheme.