Consensus For A Class Of Nonlinear Multi-agent Systems With Unknown But Bounded Measurement Noises

Consensus problem is the basis of cooperative control for multi-agent systems.Consensus means that each agent in a multi-agent system communicates according to a certain control rule and the states of all agents eventually reach agreement with the evolution of time.It has become a research hotspot during the last two decades and is widely used in the fields of robots,UAV formations and autonomous underwater submarines.In a practical system,each agent can not accurately measure its neighbor's states due to the influence of noises,time delay and communication topology,etc.It is generally considered that the state measurements of neighbors are existing disturbed quantities,in most of literature about researches for measurement noises,most of which are assumed as random variables.However,some noises are only known as belonging to a given set without any statistical properties or probability distributions,which are called unknown but bounded noises.In this paper,we study the distributed consensus control for linear and nonlinear multi-agent systems with unknown but bounded measurement noises.The main contents and contributions can be summarized as follows:Firstly,the static ?-consensus problem for a high-order linear multi-agent system is studied over a connected undirected communication topology,in which the state measurements of neighbors are affected by unknown but bounded noises.Using the dead-zone function and binomial coefficients,we propose a distributed consensus protocol.Under this protocol,all agents achieve static ?-consensus,that is,the first component of the states reaches ?-consensus,and the remaining components reach agreement at zero.Secondly,the leader-following consensus is studied for a first-order nonlinear multi-agent systems with unknown control coefficients and unknown but bounded measurement noises.Noises only affect the states measurement between followers,not between leaders and followers.Based on the condition that the communication topology contains a spanning tree with the leader as the root,undirected topology and directed topology are considered,respectively.We make use of sign function and dead-zone function for designing distributed consensus protocol to ensure that the system achieves leader-following consensus.Finally,the leader-following consensus problem is studied for a second-order nonlinear multiagent system with unknown but bounded measurement noises.The nonlinear term of the system is time-varying and related to velocity.The noises only affect the neighbor's velocity measurement,not the position measurement.Assumed that the interaction topology of the multi-agent system has a spanning tree with the leader as the root.The dead-zone function guarantees velocity consensus and the sufficient conditions for the consensus of the system are given.Under the proposed distributed consensus protocol,the system can achieve leader-following consensus.