Suboptimal MPC-based spacecraft attitude control with reaction wheel desaturation

A Time-Distributed Model Predictive Control (TDMPC) is applied to spacecraft attitude stabilization in the Local Vertical Local Horizontal (LVLH) frame with concurrent reaction wheel desaturation. This is accomplished without using thrusters, i.e., with zero fuel consumption, and exploiting gravity gradient torques. The TDMPC uses a warm-start, and a fixed number of iterations of an optimization algorithm is performed per time step to generate a suboptimal solution. An analytical estimate of the minimum number of iterations required to achieve closed-loop stability is compared to empirical estimates determined through simulations. It is shown that a small number of iterations is sufficient to perform reaction wheel desaturation maneuvers. Procedures for estimating the closed-loop regions of attraction of TDMPC are illustrated for this system.