Very Large Space Structures: Non-Linear Control and Robustness to Structural uncertainties

Continuous improvements in computational and computing sciences allow to develop more and more efficient control laws for space structures. These improvements are even more important when we are dealing with large space structures normally called VLSS (Very Large Space Structures). By virtue of this it is possible to study multi-disciplinary problems in a very detailed way, cross-linking structural, control and attitude disciplines and by providing a very meaningful case of study. In this paper a full-coupled model of a VLSS, where all the contributions coming from the flexibility effects on inertial parameters (i.e., mass, static moments, moments of inertia, modal participation factors, etc.) jointly with gravity and gravity gradient forces will be considered. This model will be used to synthetize a non-linear attitude controller named SDRE, acronym of State Dependent Riccati Equation. This enhancement of the common used linear quadratic regulator allows to have better performances of satellite attitude dynamics. This improvements is mandatory when stringent requirements on telecommunication or observation antennas pointing are imposed. Other important aspects of VLSS are relevant to the mutual interaction between attitude control and flexibility. As a matter of fact, rapid attitude maneuvers can determine great unwanted oscillations on the flexible elements of a spacecraft. In the last decades many researches, devoted to implement different devices and algorithms to undamp these oscillations have been carried out. In the present work a command shaping technique able to ”smooth” the required attitude control torque, synthesize through the SDRE algorithm, has been used to reduce the amplitude of oscillations of the solar arrays of a VLSS. Both the SDRE and the command shaping technique are based on the knowledge of the parameters (inertials and elastics) of the VLSS. Unfortunately these parameters are not always exactly known and, however they may change over the time. On account of this to explore the robustness of the proposed controller to the structural uncertainties a Monte Carlo analysis has been also performed. Numerical simulations and a critical review of the obtained results complete the work.