NMPC of Novel Quadfoil UAS for Improved Agility and Efficiency

Control techniques for lifting-body UAVs (Unmanned Aerial Vehicles) are presented that demonstrate improved efficiency and tracking through agile maneuvers and high winds. The quadfoil, a novel UAV design previously proposed and flight tested, is capable of efficient maneuvers and benefiting from crosswinds. Conventional control approaches like the previously used cascaded Proportional Integral Derivative (PID) architecture do not fully utilize the system's capability, so a Nonlinear Model Predictive Control (NMPC) approach is designed and implemented to provide local trajectory and attitude optimization in flight. Nonlinear dynamics are derived for a generic aircraft with fixed propulsors. Aerodynamics are derived from Computational Fluid Dynamics (CFD) and modeled via tensor interpolation. Wind estimation by Moving Horizon Estimation (MHE) is derived and included. The resulting performance of the NMPC quadfoil is simulated. The benefits of this control approach, which can be extended to other vehicle configurations, are discussed.