Autonomous navigation of interconnected tethered drones in a partially known environment with obstacles

Systems of tethered multicopters are multi-copter drones connected one to the other by an electric tether providing energy and communication. They are capturing the interest of researchers and industry due to their versatility and prolonged flight time. The tether enables continuous power supply from ground and fast, reliable, all-to-all communication. Nevertheless, it couples the vehicles, introducing range limitations and challenging control and navigation problems, in particular in presence of obstacles that are not fully known a-priori. To solve these challenges, a new approach based on a combination of off-line and real-time optimization is proposed. An off-line mission planning method is used to find optimal configurations for these peculiar systems in the nominal environment, guaranteeing safety with regards to the presence of the tether and obstacles. Furthermore, an on-line, real-time, reactive tracking algorithm based on Model Predictive Control is presented to bring the system to the aforementioned optimal configuration, coping with the presence of unexpected static obstacles. Such procedure also ensures collision avoidance, both for the vehicles and for the tether connecting them. The mentioned contributions are validated through simulation with realistic models of the drones and tethers.