Omni-Directional Capture for Multi-Drone Based on 3D-Voronoi Tessellation

• Published in: Drones (Basel) Vol. 7; no. 7; p. 458
• Main Authors: Cao, Kai, Chen, Yang-Quan, Gao, Song, Yan, Kun, Zhang, Jiahao, An, Di
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.07.2023
• Subjects: 3D-voronoi; Algorithms; Collaboration; Collision avoidance; dynamic target; Flow control; Methods; Motion capture; multi-drone capture; obstacle; ODMV; Robots; Strategy; Tessellation; Voronoi graphs; WPA
• ISSN: 2504-446X; 2504-446X
• DOI: 10.3390/drones7070458

This paper addresses the multi-drone formation capture in three-dimensional (3D) environments. The omni-directional minimum volume (ODMV) 3D-Voronoi diagram algorithm is proposed for the first time to achieve the two goals of (1) forming and keeping a capture and (2) planning the control action within its safe, collision region for each drone. First, we extend the traditional 2D Voronoi diagram to the 3D environment and use the non-overlapping spatial division property of 3D Voronoi diagram to inherently avoid the collision between drones. Second, we make improvements to the problem of capture angle in our minimum area strategy and propose an omni-directional minimum volume strategy to accomplish the effective capture of a target by constraining the capture angle. Finally, the wolf pack algorithm (WPA) with variable step size is introduced to provide a movement strategy for multi-drone formations. Thus, the proposed ODMV can also achieve dynamic target and multi target capture in environments with obstacles. The Optitrack motion capture system and Crazyflie drones are used to conduct the multi-drone capture experiment. Both simulation and experimental results are included to demonstrated the effectiveness of the proposed ODMV method.