The Asynchronous Containment Control Problem Of Multi-agent Systems Under DoS Attacks

The research of multi-agent systems is an important field of control science.By designing appropriate control protocols,agents exchange information with each other through a communication network.The agents run as a whole and the system executes challenging tasks with higher efficiency,flexibility and robustness.Multi-agent systems are widely used in various fields in the world,such as smart grid,UAV swarm control,multi-target navigation and tracking,etc.Based on the theoretical research on the consensus of synchronous multi-agents,researchers find out that the synchronous updating of agents is difficult to be achieved due to technical limitations in actual engineering applications.Hence,the asynchronous multi-agent systems are proposed which allows each agent to update its state at its own clock.Moreover,there exists uncertain interference in practical problems,such as Denial of Services(Do S).In this thesis,effective proofs are given for the containment control problems derived from second-order asynchronous multi-agent systems and Do S attacks.The main research content is summarized as follows:1.The asynchronous containment control problem of second-order multi-agent systems is studied.Based on the communication topology and control protocol under asynchronous conditions,the multi-agent system is transformed into a new system to describe the state between leaders and followers and the property of the stochastic matrix and scrambling matrix is used to prove the consensus of the system.Based on the infinite SIA matrix product convergence property and definition of convex hull,the containment problem of second-order asynchronous multi-agent systems is addressed.Finally,the effectiveness of the theory can be verified through numerical simulation.2.The asynchronous containment control problem of second-order multi-agent systems under Do S jamming attacks is studied.A specific control protocol under Do S attacks is designed and a gain coefficient is to be determined.Based on matrix theory,the multiagent system is transformed into an error system.Firstly,the binary relationship is used to prove that there is a directed spanning tree in the asynchronous communication topology.Based on the non-negative matrix theory and graph theory,the range of gain coefficients that can address containment control problem is obtained.Finally,a necessary and sufficient condition about the asynchronous containment control problem of the second-order multi-agent system under Do S attacks is proposed.The numerical simulation is presented to prove the effectiveness of the result.3.The asynchronous containment control problem of the second-order multi-agent systems under Do S attacks in noisy communication is studied.Due to the exogeneous white-noise disturbance,each agent has difficulty updating its state accurately.By designing an anchor unit inside the convex hull conducted by the leaders,the error system about the follower state and the anchor unit is constructed.Based on the Lyapunov theory and matrix theory,a necessary and sufficient condition is presented.Especially,the conclusion of the problem is also feasible under white-noise disturbance.Finally,the effectiveness of the theory can be verified through numerical simulation.