Algorithm for energy efficient inter-UAV collision avoidance

摘要

Unmanned Aerial Vehicles (UAV) are fast gaining popularity. Due to their many desired features, in the near future, UAVs will be an inevitable part of many fields. This increased use of UAVs, has given rise to the idea of multi-UAV systems, commonly known as UAV swarms, where a number of UAVs ranging from two to hundreds have to collaborate with each other and perform a common shared task or tasks. When sharing a common air space and flying in close proximity of each other, inter-UAV collision avoidance becomes an important factor in UAV swarms. Thus, for effective use of UAV swarms, it is essential to have an efficient inter-UAV collision avoidance mechanism. Although there are a number of suggested techniques, avoiding inter-UAV collisions while minimising the energy consumption of the UAV system is a challenge. In this paper, a Potential Field Method based algorithm to avoid collisions of a UAV system, considering the minimising of system energy usage, is suggested. We consider the energy consumption of a UAV system to depend on the distance travelled and the turns - angular changes, in trajectories. Our method reduces the angular changes in trajectories by an average of 36% and total travel distance by 6-8%. The total time taken to achieve targets is reduced by 3-14% depending on the scenario.