Swarm Intelligence based Collision Avoidance Between Realistically Modelled UAV Clusters

In this paper problem of collision avoidance in group of UAVs is addressed. Algorithms for two types of collision avoidance: Group-wise and Individual are successfully developed in the three-dimensional plane using various basic swarming behaviors. Limit on both horizontal and vertical acceleration is applied to cater the hardware constraints of UAV. Effect of these basic laws has also been studied on the group behavior of swarms. It is shown that when cohesion rule is applied an equilibrium condition is reached in which all the UAVs settles down at same height on a circle of constant radius. It is also shown analytically that this equilibrium condition is stable for all values of velocity V and acceleration constant k. Finally, noise of zero mean and standard deviation sigma is added to the position of UAVs to study the effect of position uncertainty on group behavior and on both types of collision avoidance.