Formation Circling Control Of Detect-pursuing Structure With An Uncertain Dynamic Target

With the development of communication,computer and artificial intelligence technology,as well as the deepening of task scene complexity,the formation control of multi-agent system has become a hot research field.The formation control of multi-agent requires that the state of agents should be consistent,and the system should form a certain formation for specific interested targets.At present,the following problems exist in the measurement based formation enclosing control: the situation that some agents cannot measure the relative position information with the target is ignored.In this case,the existing formation enclosing controller is difficult to meet the requirements of formation cyclic.In order to solve the above problems,we introduce a detecting-pursuing structure among agents,and design a multi-agent formation enclosing control strategy based on this.The main contents and innovations of this paper are as follows:1.Formation enclosing control based on cooperative estimation.We first introduce the detecting-pursuing structure,and divide agents into detector and pursuer.The detector uses the adaptive estimation method to estimate the unknown state of the target,and designs the control laws of normal direction(direction to the target)and tangent direction(direction perpendicular to the normal)to achieve the tracking and enclosing of the target.For the pursuer’s control law,we also decouple it into two directions: normal(pointing to its upper navigator)and tangent,and complete the distance tracking and relative angle tracking of the leader respectively.By using Lyapunov stability theory and La Salle invariant set theory,the asymptotic stability of the system is proved,and the correctness of the theory is verified by numerical simulation.2.Formation enclosing control based on uncertain state estimation.Different from the research content one,which only considers the unknown state of the target,the estimation of the upper navigator’s state of the pursuer is also included here.For the detector,the design of the controller is similar to the content one,which increases controller in the tangent direction to keep the relative neighbor angle,and realizes the formation between the detectors.For the pursuer,it is necessary to design a cooperative adaptive estimator to estimate the unknown state of the navigator,and then design the pursuer’s normal and tangent control according to the design method of content one.By using Lyapunov stability theory and La Salle invariant set principle,the asymptotic stability of the system is proved,and the correctness of the theory is verified by numerical simulation3.Design of simulation platform for multi-agent control experiment.Based on the open source ground station QGround Control,a multi-agent control scheme is designed.By adding the check function,the ground station can transmit the status information of all selected vehicles or temporarily joined vehicles at the same time.Then,the receiving and sending interface function is designed to realize the formation flying in the leader-follower mode.It provides a software verification platform for multi-agent formation enclosing control algorithm.