Research On Multi-agent Leader-following Formation Control Based On Bearing

At present,the multi-agent cooperative control is a hot topic of current research,among which the multi-agent consensus control and the formation control are two fields receiving much focus.Considering that the formation control is widely used in many fields such as aviation,drone formation,agriculture,etc.,we mainly address the leader-following formation control of multi-agent systems in this thesis.In formation control,there are three basic controls: rotation,scaling,and translation.Researchers often use distance rigid graph theory for formation control i.e.,the relative position or relative distance constraint between target formation agents.Due to the invariance of relative positional constraints between the agents during the formation translation movement,the relative position based control can be applied to the translational formation maneuver control;Due to the invariance of relative distance between the agents during the formation rotation and translation,the distance-based control is often used to implement translation and rotation control of formations.Recently,more and more researches have had great interests in bearingbased formation control.This stems from the fact that in the bearing-based formation process,the target formation is determined by the bearing constraints.Since the bearing constraints between the agents in the formation translation and scaling control are invariant,the bearingbased control can be used to implement the translation and scaling operations of the formation.Considering the advantages based on orientation,this paper adopts the bearing rigid graph theory to study the leader-following formation control.Although the typical first-order and second-order intelligent body dynamics models can simplify the research of multi-agent systems,they can't reflect authentically the dynamics of the intelligent body in engineering practice.Therefore,this paper adopts a third-order system as the plant model to study the leader-following formation control.The paper first introduces the research status of multi-agent cooperative control,and Secondly,the mathematical model of agent in studied leader-following formation control problem is devised.The leader-following formation control problem is transformed to the problem that the formation needs to be converged to target formation.The target formation is determined by the ideal bearing constraint and scaling factor between the agents.Combined with the research in literature,it is derived that the target leader-following formation has unique conditions.Thirdly,the control law based on the relative position,velocity,acceleration between the first leader and the second leader and the second leader's model.This can to maintain the relative position between first leader and second leader as a constant value.The stability proof is carried out by Routh-Hurwitz.Fourthly,the distributed control law of the followers is designed,in order to achieve the ideal bearing between the agents.Finally,the numerical verification of the leader-following formation is performed by numerical software.