Research On Consensus Algorithm For Multi-Agent Systems Based On Dynamic Interactive Topology

In recent years,with the rapid development and integration of computer network technology,information technology,automation technology and artificial intelligence technology,distributed cooperative control of multi-agent systems has become a very important and hot research field.It has attracted great attention from researchers in the fields of mathematics,control,physics,biology,robotics,computers and artificial intelligence.Among the research directions of multi-agent systems,consensus is one of the most fundamental and important research directions.However,in practical applications,because of the large scale of the multi-agent systems,the existence of external interference makes the communication connection between the agents unreliable,the scope of the communication and sensing of the individual unit of the agent is limited,and the deterioration of the external working environment may also cause the failure of the transmission channel itself.As a result,communication topology among multi-agent systems is not always fixed.In addition,considering the limitations of the resources,calculations and communication capabilities of each agent itself,event-triggered control can better alleviate the communication burden and effectively save communication resources than traditional time control.On the basis of summing up the previous works,this paper takes the consensus problem of multi-agent systems as the background,combining with the method of average dwell time,and using the output feedback with unknown system state in turn,corresponding consensus control algorithms are designed for nonlinear multi-agent systems with intermittent communication and switching topology,respectively.Finally,based on the event-triggered technology,a distributed consensus protocol is designed,which can effectively reduce the update frequency of the controller.The main contents of this thesis are as follows:First,the output information of agents can only be measured and utilized is considered,and a novel consensus controller protocol based on state observer is proposed for the nonlinear multi-agent systems with intermittent communication.At the same time,an algorithm is designed to select the corresponding controller gain matrix and coupling strength coefficient,and two optimization problems are given to optimize the related parameters.By using the tools of matrix analysis and stability theory,it is theoretically proved that the designed controller can realize the consensus tracking of nonlinear multi-agent systems with intermittent communication,and the effectiveness of our proposed method is verified by the simulation of a practical example.Next,the nonlinear multi-agent system with switching topology is considered,in which the multi-agent system is composed of leader layer agents and follower layer agents.On the basis of the previous research,we consider that the global information of the system is difficult to obtain,and the adaptive parameters are introduced to improve the self-tuning ability of the whole system.By using the state information of the agent itself and the output information of its neighbor agents,a completely distributed observer-based consensus protocol is designed.Through rigorous theoretical analysis,it is proved that the controller designed can make the state of the follower layer agent track the state of the corresponding agent in the leader layer,at the same time,the simulation of the practical example effectively illustrates the feasibility of our research method.Then,combined with the event-triggered controller communication scheme,by using the state information of the agent distributed event-triggered consensus protocol is designed for nonlinear multi-agent systems with intermittent communication.In order to get the maximum communication recovery time,we give an optimization problem to get the optimal value for the time interval parameter of intermittent communication.Compared with the existing literature,we first introduce the event triggering mechanism into the consensus problem of multi-agent systems with general directed dynamic interactive topology,and prove that the Zeno phenomenon can be avoided,and the effectiveness of the proposed method is verified by the simulation of the practical model parameters.Some conclusions are presented at the end of this thesis and the further research direction is discussed.