Nonlinear Incremental Robust Model Predictive Control for a Nonlinear Aeroservoelastic System

ifasd2024 Tracking Number 181

This paper presents a novel approach to suppress gust-induced limit cycle oscillations (LCOs). The proposed method integrates incremental nonlinear dynamic inversion (INDI) and robust nonlinear model predictive control (NMPC) with tightened constraints. The INDI method estimates and actively rejects gusts, resulting in a reduced disturbance residue. The upper bound of the disturbance residue can be estimated either online or offline and is used to bound the maximum state deviation caused by uncompensated disturbances, thereby imposing tightened constraints for the NMPC scheme to improve the robustness of constraint satisfaction and stabilize the system. Simulation results on a 2-D nonlinear aeroservoelastic wing demonstrate that the proposed method stabilizes the nonlinear system and reduces the disturbed peak plunge and pitch motions by up to 20.69% and 33.70%, respectively. Additionally, the method mitigates the conservativeness of the robust NMPC, with an 81.67% reduction in the offline estimated upper bound of the disturbance residue. The online estimation of the disturbance residue captures its peak value while further relaxing the tightened constraint set when disturbance effects are small. The proposed control scheme effectively suppresses gust-induced LCO motions and reduces the conservativeness of the tightened constraint sets used in the robust NMPC.