Research On Consensus-based Grouping For Homogeneous UAV Swarm

With the breakthrough of related technologies,drones have become more efficient and intelligent.UAVs are able to perform tasks autonomously and search,identify,track and strike targets,greatly improving the efficiency and reducing the cost.The UAV swarm will coordinate the control of a certain amount of the drones as a whole.Compared to a single drone,a drone swarm can take advantage of scale,but it also brings control problems.Grouping the UAV swarms can effectively solve the difficulties caused by the increase in the size of the swarm.At present,the research on grouping problem of UAV swarms is still insufficient.It is important to realize the adaptive grouping of UAV swarms to improve the efficiency of control.This paper studies the grouping problem of UAV swarms.Mainly for the selforganization,high dynamic and large-scale distributed characteristics of UAV swarms,the grouping method based on distributed consensus is explored,which involves technical problems such as swarm grouping,group leader election,group member change and fault recovery.The main research results are as follows:(1)Analyzed the grouping problem of the UAV swarm and modeled it.By combining the working principle and process of the UAV,the control method of the UAV swarm,the related research of the distributed consensus,and based on certain simplifications and assumptions,the UAV swarm grouping problem model is established.Single drone of the swarm and overall flight control rules are set.(2)The consensus-based adaptive UAV swarm grouping method is designed,including four key algorithms: swarm grouping,group leader election,group member change and fault recovery.The above consensus-based method realizes the initial grouping of swarms,performs leader elections in the drone group,controls the drone groups to merge,supplements the isolated drones,and takes different measures to deal with group leader or group member failures.Satisfying the requirements for adaptive grouping of UAV swarms also improves the efficiency and robustness of UAV swarm control.(3)The simulation experiment was designed with reference to the actual UAV swarm.The reasonable evaluation indicators were used to compare and analyze the swarm grouping,group leader election,group member change and fault recovery algorithms proposed in this paper.The results prove the effectiveness of the consensusbased grouping method designed in this paper in the case of homogeneous drone swarms.