Self-Organization Algorithms for Movement Control of a Multicopter Swarm

The success of global satellite navigation systems has enabled the widespread use of low-cost multicopters (MCs) across various fields. While individual MC usage is now commonplace, designing and operating large groups of MCs with a unified goal presents a new challenge. This article addresses a key challenge in implementing the behavior of a so-called MC swarm, which is a dense group of multicopter agents. It focuses on developing control algorithms for individual MCs and defining the informational interactions between them. These interactions ensure tasks such as swarm self-organization and coordinated movement without disrupting the swarm's structure. The management strategy relies on virtual physical principles of attraction and repulsion, where the forces' magnitude depends on the distances between objects. The developed algorithms enable autonomous control of the MCs and are designed for onboard implementation. While the algorithms are identical for all MCs, they can produce different individual behaviors to maintain group integrity. The article describes algorithms that ensure the following: self-organization of the swarm around a set concentration point at arbitrary initial coordinates of the MC; complex movement, represented as a combination of the movement of a central point (the pole) and the rotation of the swarm around it. In any scenario, the collision prevention algorithms for MCs operate continuously. The effectiveness of the algorithms is demonstrated through computer simulations, providing both qualitative and quantitative performance assessments.