Research On Distributed Consensus For Multi-agent System With State Constraints

In recent years,a siginificant amount of research efforts have been focused on the coordination control of multi-agent systems in various areas including computer engineering,natural science,and social science.The main research directions of multi-agent systems can be divided into consensus,tracking,formation and containment.As a basic problem in distributed cooperative control for multi-agent systems,consensus has been extensively studied during the last decade.Most of the existing consensus research assume that the states of the agent can vary without any constraints.In practice,the multi-agent systems need to reach a consensus while the state of each agent must lie in a certain convex constraint set.In this paper,we study the distributed conensus problem for multi-agent systems with state constraints and uncertainties.The main results are as follows:For the multi-agent systems with state constraints under a connected and undirected communiction topology,we study the distributed consensus problem in this situation.The state constraints for each agent can be indicated as individual closed convex sets and the intersection of these sets is not empty.When the multi-agent systems reach consensus,the consensus point lie in the intersection.In this thesis,based on the distributed consensus algorithms,we utlize the characteristic of projected operator to make sure each agent presrve in state constraints.Consider the exist of parametric uncertainties,we proposed a distributed adaptive control algorithm to achieve consensus.Next,we extend the result to double integrators system,and using Lyapunov stability theory to make sure the system is stabilized.When the communication topology is directed and contain a directed spanning tree,we study the distributed consensus alogrithms for multi-agent systems with time-varying state constraints and uncertainties.We convert the multiagent system with state constraints into a new one with no constraints.The stabilization and consensus of transformed states can ensure the consensus of original multi-agent combinging with the predefined time-varying state constraints.Since the unbalance of Laplicaian matrix of the directed topology and the system with uncertainties,we use model reference adaptive control to analize the problem and we proposed an adaptive contorl algorithm for the system with uncertainties.Morever,we study the system with time-varying state constraints and prescribed performance.The consensus convergence are provided via Lyapunov stability and input-to-state stability.