Deterministic Reconfiguration of Flight Control Systems for Multirotor UAV Package Delivery

Abstract

Multirotor UAS are prime candidates for autonomous package delivery due to their VTOL capability and payload carrying capacity. The effect of payloads on flight control system performance is investigated for three different inner-loop flight control system architectures, namely, explicit model following, nonlinear dynamic inversion, and incremental nonlinear dynamic inversion. Outer-loop flight control systems are wrapped around the various inner-loop architectures for waypoint tracking control. The flight control systems are designed and optimized using CONDUIT R to meet a common, comprehensive set of stability and performance specifications. Deterministic reconfiguration was designed for each inner-loop control architecture to account for the change in vehicle dynamics when a payload is added or removed. Robustness analyses are conducted considering both deterministic payload variations and modeling uncertainty. A notional package delivery mission scenario is simulated using a full-flight envelope stitched model with measurement noise and turbulence models identified from flight test data. The mission scenario is simulated for three different cases to evaluate the baseline performance, the degraded performance when a payload is added, and the recovery of performance with deterministic reconfiguration of the flight control systems.