NONLINEAR ROBUST MODEL PREDICTIVE CONTROL FOR LIFTING BODY RE-ENTRY FLIGHT ATTITUDE CONTROL

Abstract

The dynamics of re-entry vehicles are time varying and notoriously nonlinear due to the large flight envelope and nonlinear aerodynamic flow phenomena. Accuracy of guidance and control of the re-entry vehicle along a predefined trajectory is crucial because of operational and safety considerations. This paper presents the design of a nonlinear attitude controller based on a Robust Model Predictive Controller (RMPC) combined with Feedback Linearization of the vehicle dynamics. Since the re-entry vehicle model contains uncertainties, FBL will not fully linearize the system. This, combined with its constraint handling capabilities, is the motivation for the application of RMPC. The uncertainty is modeled as perturbations of the aerodynamic parameters within a given percentage of their nominal value. The designed controller was tested using the GESARED simulation toolbox and shows excellent performance. The state and input constraints are satisfied, and the robustness of the controller to variations in the aerodynamic parameters has increased significantly compared to a nominal, non-robust controller.