Mutual collision avoidance in the formation of the unmanned aerial vehicles

A group flight of hypothetical unmanned aerial vehicles is considered, during which mutual collisions are possible. A collision avoidance strategy is proposed during the group flight formation of these vehicles. This strategy is based on the method of artificial potential field. The method consists in that when a collision risk is detected between agents, control signals for the longitudinal, lateral channel and thrust channel are automatically generated in the control system of each agent and the control objects reach a safe distance relative to each other. Collision avoidance algorithms have been simulated. At the first stage, collision avoidance algorithms for each channel are simulated separately. Agents move at specified coordinates relative to the virtual leader. Then two UAVs are selected, the coordinates of which differ only on one channel, and one of the agents deviates towards the other. Then collision avoidance algorithms are simulated for all channels simultaneously. Three agents are considered, which move along given coordinates relative to the virtual leader, thus forming a system. Then the second and third agents deviate towards the first. Thus the risk of UAV collision arises. The simulation results demonstrate that these control algorithms prevent all possible collisions.