Research On Containment Control Problems Of Multi-agent System With Switching Topology

Multi-agent system is a group which consisting of several simple individuals and the cooperative control of multi-agent system mainly refers to that all individuals can complete assigned tasks by local information exchange,such as the formation of multi-aircraft,the cooperation of multi-motor and the cooperative handling of multi-robot are typical cooperative control problems.As the widespread existence of such systems in nature and real life,it attracts many scholars in the field of control,physics and other fields,and has achieved rich results.Where,containment control is a basic problem in the field of multi-agent system,which purpose is that all followers can eventually converge to the convex spanned by leaders via designing an appropriate control protocol based on individuals.This dissertation mainly investigative the containment control with switching communication topology.The main contents are as follows:1.The containment control of first-order multi-agent system with random switching topology is investigated.Consider the continuous multi-agent system with multiple static leaders and followers.Based on the designed control protocol,the containment control problem can be solved when the union graph of system is globally reachable,and the simulation results verify the validity of the results.2.The containment control of second-order multi-agent system with random switching topology is studied.The discrete multi-agent system with multiple dynamic leaders and followers has been considered.Based on the designed control protocol,by constructing a mean square error function and utilizing equivalent transformation of the system,the second-order containment control problem is transformed into a reduced order system stability problem.The sufficient condition for the stability of the error system is obtained.Under the sufficient condition,the containment control problem can be solved.The simulation results verify the validity of the results.3.In the case of switching communication topology,a class of surrounding control of multi-agent system is considered.In the two-dimensional plane,consider a multi-agent system consists of one static leader and multiple followers,the purpose is that all followers can eventually converge to the circle centered by the static leader and surround in the circle in an uniform way.To this end,a distributed control protocol is proposed for the follower to estimate the static leader position,and the sufficient condition for all the followers can converge to the specified circle is obtained.Secondly,a distributed control protocol is designed for followers to move around the circle.All followers can surround in the circle in an uniform way if the sufficient conditions is satisfied.Simulation results verify the effectiveness of the proposed method.