Research On Formation Control And Task Planning Algorithm Of UAV Swarm

Unmanned Aerial Vehicle(UAV)formation has been applied more and more widely in many fields in recent years with many advantages such as good flexibility and strong maneuverability.UAV formation control needs to ensure formation formation,maintenance and obstacle avoidance.If inappropriate control methods are adopted,there will be many problems,such as the oscillation of target position,the problem of target unreachable,the unsafety of obstacle avoidance and so on.Task planning of UAV formation requires assignment of task set and decision of task execution sequence for each UAV.In actual situations,there are differences in resource requirements among various missions and operational capabilities among various UAVs,and the influence of task execution sequence of UAVs should also be considered.All the above problems need to be considered and solved in the research process.This paper takes UAV formation as the main research object and studies formation control under dynamic obstacle environment and multi-type task assignment of heterogeneous UAV formation.Firstly,the formation control of UAV based on artificial potential field is studied.In order to solve this problem,an improved artificial potential field function,including formation generation and maintenance potential field function and dynamic obstacle avoidance potential field function,was designed by combining the ideas of virtual structure method and leader-follower method and considering the vector field of threedimensional space.In the potential field function,the target position attractive force is used to lead the UAV to form a formation without shaking,so as to avoid the problem of oscillation at the target position.The steering force is designed to change the speed direction of the UAV to avoid the local minimum problem.The restorative force is designed to pull the UAV back to the preset trajectory,and a deceleration strategy is added to avoid the problem of oscillation around the preset trajectory.Based on the above function model,a dynamic obstacle avoidance algorithm for a single UAV is proposed to generate a specific obstacle avoidance trajectory.The formation and maintenance of UAV formation and dynamic obstacle avoidance are studied.The formation control algorithm of UAV based on artificial potential field function is proposed to avoid obstacles in the environment of dynamic obstacles,and the formation can continue to form and maintain after the completion of obstacle avoidance.Simulation results show that the proposed algorithm has advantages in performance in the scenario.Finally,the task assignment problem of UAV formation based on auction mechanism is studied.According to the different resource requirements between tasks,the different execution capabilities between UAVs and the impact of task execution sequence,sevral cost functions and a multi-type task assignment model are designed,and the goal of the model is to maximize the global task execution benefit.This benefit is related to the maximum flight range and limited execution capability of the UAV,as well as the resource requirements and time window constraints of the mission.Based on the above model,a task assignment algorithm based on auction mechanism is proposed to generate the conflict-free task execution sequence of each UAV.Simulation results show that the proposed algorithm has advantages in planning path length,execution time and energy consumption.