ESA GNC Conference Papers Repository

Linear modeling and robust control of a flexible space manipulator
B.D.D. Dellandrea, A.P.G. Perez Gonzalez, V.D. Dubanchet
Presented at:
Salzburg 2017
Full paper:

This article aims at presenting the controller of a flexible robotic arm mounted on a servicer satellite. The synthesis is based on robust control techniques for ensuring stability against variations of the nominal configuration of the robotic arm, such as mass, inertia, relative angles between the segments or disturbances at hub and joints level. The final aim of this article is to present the robustness of a controller against these parametric uncertainties. First, the linear dynamic model is derived for the system composed of a rigid hub and a flexible robotic arm. The modeling technique is based on [1], called the TITOP (Two-Input Two-Output Port) modeling technique, which allows the interconnection of flexible substructures in chain-like assembly considering revolute joint between in the robotic arm segments. The modeling technique provides a LFR (Linear Fractional Representation) model of the whole system which could take into account uncertainties in the flexible modes, damping coefficients, relative angles between the arm segments, hub mass and inertia, and other design parameters. Second, the LFR model of the system is used to synthesize a robust controller in the sense of the H-infinity norm. The system is controlled by actuators located at the hub, which can rotate and translate around its center of mass. Mu-analysis is used to find the upper and lower bounds of the uncertainties, finding the ones which are more likely to destabilize the system in closed loop . Finally, simulations are run for different values of the system's parameters in order to validate the expected boundaries for robust performance. Conclusions about the controller tuning and the system modeling are stated in order to find future improvements of this approach.