Balázs Németh, Attila Lelkó, Zoltán Antal, Ajtony Csaba, Collision-free trajectory design for dance choreography of virtual drones in hierarchical structure, IFAC-PapersOnLine, Volume 55, Issue 6, 2022, Pages 502-507, ISSN 2405-8963.
This paper proposes a collision-free trajectory design for dance choreography of virtual drones using a hierarchical design structure. The trajectory on two levels is designed, i.e., on the level of individual virtual drones and on the level of their centralized coordination. The design on the drone level through reinforcement learning has been carried out, with which dance steps of virtual drones in their motion can be achieved. The coordination level contains a prediction algorithm, which guarantees the avoidance of collision, even at the faultiness of some virtual drones. The effectiveness of the hierarchical design on the example of a Viennese waltz dance choreography in augmented reality has been illustrated.