Scalar Reference Governor Bank for loosely cross-coupled MIMO systems: application to Satellite Attitude Control

The control strategy of complex engineering systems, e.g., in automotive, aeronautical and space applications, are oftentimes well-established and there is limited space to integrate major novelties in the design of control laws. As a consequence, the problem of managing the system constraints is addressed with overconservative and sub-optimal solutions. With a particular focus on satellite missions, this paper proposes a Reference Governor-based approach to design the optimal safe trajectory so as to exploit the full capabilities of an established closed-loop Multi-Input Multi-Output (MIMO) system subject to state and input constraints. While not interfering with its stability properties, the Governor predicts the future evolution of the closed-loop system and modifies the reference to track in case constraints are at risk. The computationally attractive Scalar Reference Governor is compared to the Vector Reference Governor, which is optimal for MIMO systems. Finally, a sub-optimal fast Governor is proposed for MIMO systems with limited coupling. Numerical simulations are run on the CNES high-fidelity simulator developed for the Microcarb mission and illustrate the advantages of the proposed methodology.