Research On Multi-UAV Cooperative Target Hunting

With the maturity of UAV technology,coordinated operations of multiple UAVs that can independently perform various tasks have great development prospects in the military field.Aiming at the combat scenario of multi-UAV hunting ground moving targets,this article has designed a set of multi-aircraft cooperative target hunting system with a complete combat process,completed the algorithm theory research of the decisionmaking layer,multi-UAV system architecture construction and Gazebo simulation verification,and finally carried out the flight verification on the multi-UAV platform.First of all,this paper studies the UAV route planning method for multi-aircraft cooperative target hunting in two aspects: multi-aircraft coordination and task execution.In terms of multi-aircraft coordination,the UAVs need to reach the target area together in formation,and there is a risk of collision between UAVs.To solve this problem,the collision avoidance and obstacle avoidance algorithm based on artificial potential field is studied,and the algorithm is improved to adapt to the dynamic environment.The improved algorithm can not only avoid obstacles,but also realize the tracking of moving targets.At the same time,the formation was designed combined with the leader-follower formation strategy.After that,numerical simulation was used to verify the ability of multiple drones to quickly form the required formation.In terms of task execution,it includes two parts: collaborative search and target hunting.Among them,the core of the search trajectory planning algorithm is polygonal runway search and target digital search.To quickly find the target number board,the concept of the best observation point is introduced while searching for the target number board.The target hunting algorithm adopts the multi-aircraft phase separation method to realize the tracking and hunting of the target.In the end,multiple drones can quickly find the target and hunt the target.Secondly,in view of the task allocation problem of multi-UAV to multi-target,the distributed task allocation algorithm based on the contract network protocol is studied.Through the four phases of "target finder issuing bids-bidders bidding-bid evaluationassigning tasks",the task allocation is completed.Aiming at the task scenario,this paper adopts the "3+2" target hunting strategy to design a task allocation algorithm that can redistribute different task types multiple times.Finally the problem of redistributing multiple task types(cooperative search,target tracking)during the hunting process is solved.After using numerical simulation tools to verify the allocation effect of 10 UAVs hunting 4 targets,the effectiveness of the allocation algorithm was proved.Finally,taking full account of the complexity of the target-hunting scene,this paper designs a systematic collaborative simulation program to complete the simulation of the combat process including the UAVs starting from the landing area,flying in formation to the target area,searching for the target,assigning the UAV to search for target numbers,and target hunting.This simulation program’s further verification was completed in the Gazebo environment.In addition,three quad-rotor UAVs were used as the flight platform,and the actual flight verification of three UAVs against two targets was completed.In the end,the UAV successfully completed formation assembly,coordinated search,target tracking,and formation return.Throughout the combat process,the effectiveness of the multi-UAV collaborative hunting multi-target algorithm studied in this paper is verified.