Fast Formation of Swarm of UAVs in Congested Urban Environment

As Unmanned Aerial Vehicles (UAVs) become more readily available and reduce in cost, using multiple UAVs simultaneously to accomplish a task becomes increasingly attractive. Once swarms of UAVs share the same workspace (operational environment) it is necessary to have a means to rapidly adopt an optimal collision-free formation in the workspace. A popular approach for formation of an optimal swarm is to use Particle Swarm Optimization (PSO) techniques. A variant of PSO was recently introduced called nPSO which claims to exhibit more rapid convergence than other variants. In this paper nPSO is applied to the problem of finding optimal positions of UAVs forming a swarm in presence of large obstacles such as buildings in an urban environment. The experiments show that no more than 1000 iterations are required to obtain near optimal formation of swarm of UAVs for different maps, including maps relating to congested environments.