Distributed MPC method for collaborative aerial transportation: from design to experimental validation

Stabilizing a load connected through a cable to a single UAV poses a complex challenge, constrained by payload weight restrictions. Consequently, collaborative aerial transportation has aroused the interest of numerous research projects. In this work, we introduce a Distributed Nonlinear Model Predictive Control (NMPC) framework designed to regulate the complete pose of a cable-suspended load. It leverages on a partition based approach and on the kinematic structure of such systems, achieving good tracking performances with agile trajectories while being consistent with some safety-related constraints. The advantages of this state-of-the-art controller, with respect to previous works concerning Centralized NMPC for collaborative aerial transportation, is the lower computational burden and the scalability with an increasing number of robots. Simulations are undertaken to affirm the algorithm's viability, employing a simplified yet comprehensive model. \\ The study focuses also on the experimental direction showcasing the process of constructing a functional Fly-Crane system, comprising 3 drones and a platform, ready for real experiments. Furthermore, experimental results demonstrating the application of a centralized NMPC to the Fly-Crane system are presented.