Adaptive Control Design for Multi-UAV Cooperative Lift Systems

Abstract

Multi-unmanned aerial vehicle (UAV) cooperative lift systems use multiple UAVs to collectively lift and transport payloads. These systems have unique benefits over standard single-vehicle logistics paradigms in that they distribute lift capacity among several potentially inexpensive and man-portable aircraft, and furthermore provide redundancy to guard against vehicle failure. However, several challenges arise when multiple vehicles are tasked with coordinating stabilization and control tasks for an unknown payload, particularly when there is also uncertainty regarding relative vehicle placement and orientation. This paper presents a novel adaptive control framework for modular cooperative lift systems. The algorithm uses an extended Kalman filter to update system parameters within a control allocation scheme. Thrust vectoring is employed to ensure adequate yaw control authority, and null-space excitation is used to speed estimator convergence. The adaptive control approach is designed to execute a full payload transportation mission starting from rest on the ground. Extensive simulation studies and preliminary flight experiments evaluate robustness of the adaptive control scheme and explore tradeoffs with respect to several system parameters and levels of uncertainty. Overall, the proposed control and estimator scheme is shown to be highly effective in stabilizing multi-UAV payload systems under realistic uncertainty in vehicle configuration and payload parameters.