LOCAL PURSUIT STRATEGY-INSPIRED COOPERATIVE FORMATION FLIGHT AND COLLISION AVOIDANCE FOR UAV CLUSTER

The unmanned aerial vehicles (UAV) cluster, compared to a single UAV, has better application potential in implementing complex and dangerous tasks. In recent research, cooperative formation control strategy and obstacle avoidance have always been focused to secure the safety of flight and task implementation. In this research, we propose the Local Pursuit (LP) strategy-inspired cooperative flight and collision avoidance model inspired by foraging ants, which consists of a formation control module and an obstacle avoidance module. In the formation control module, under the guidance of LP strategy, the follower UAVs will always point their velocity towards the leader UAV (which can be virtual) like ants following the ant ahead. Meanwhile, the velocity and formation position of each follower UAV can be adjusted by the 'speed control parameter (SCP)' related to the speed of the follower and the constant bias between the leader and the follower which is determined by the desired formation. Considering there are cases that some areas threatening the safety of the UAV need to be avoided, which can be regarded as external obstacles, an obstacle avoidance module based on velocity alignment with "shill" agents (virtual UAVs moving outward from obstacles to real UAV) is added for effective obstacle avoidance. The proposed method has limited parameters that need to be tuned. Finally, to show the efficiency of the proposed method, the simulations of UAV cluster formation flight and obstacle avoidance by using the proposed method are conducted through MATLAB.