Performance and Control of Multirotors Formation with Sling Load in an Urban Environment

In recent years, autonomous delivery has garnered considerable attention, a market with substantial growth potential. Using small multirotors for door-To-door delivery offers a number of advantages, including reduced shipping costs and delivery times, as well as high reliability and convenience. A significant drawback is, however, that the currentlyoperational multi-rotors have relatively limited lifting capacity. This can be mitigated, to some extent, by utilizing a formation of UAVs to accomplish the same task. Developing control algorithms for such systems that carry a sling load can be challenging due to the complex dynamics of the formation, as well as the complicated aerodynamics of the rotors. A simplified aerodynamic model, in which the force generated by a rotor is determined solely by its rotational speed, is usually adopted for the controller design. This study introduces a more realistic aerodynamic model to evaluate the validity of those common practices. It demonstrates that the commonly used model is inadequate in windy conditions, especially in urban environments with high wind gradients. By exposing a formation of four multi-rotor UAVs carrying a sling load to turbulent winds, it is shown that a trajectory tracking guidance logic with a force controller, designed using the simplistic model, can even crash. As part of a revised controller scheme, system performance can be markedly improved by implementing an approximation model based on gain scheduling for a range of flow velocity conditions.