Adaptive neural network based fixed-time attitude tracking control of spacecraft considering input saturation

Abstract

Aiming at the issues of actuator saturation, inertia uncertainties, and external unknown disturbances in the attitude tracking control process of spacecraft, an adaptive fixed-time attitude control method is proposed, which is based on a radial basis function neural network (RBFNN). Firstly, a spacecraft attitude kinematics and dynamics model is established based on the quaternion method and a Gaussian error function is introduced to constrain the controller amplitude. Secondly, the external unknown disturbances are addressed by a fixed-time disturbance observer, and the controller is designed utilizing the backstepping method. To eliminate the adverse effects caused by actuator saturation, we design an enhanced auxiliary system to improve the stability of the system. Aiming at inertia uncertainties, RBFNN is used to approximate it, and an innovative fixed-time convergence adaptive law with RBFNN weights is devised. Subsequently, based on Lyapunov theory, the fixed time stability of the closed loop system is proven, and an expression for the settling time is given. Finally, simulation analysis validates the effectiveness of the designed controller.