Stochastic Consensus Of Multi-Agent Systems With Measurement Noise And Communication Delay

Based on the control theory of linear systems,the distributed consensus of multi-agent systems with communication interference is studied in this thesis.This thesis is mainly divided into three parts.In the first part,an asymptotically unbiased mean square linear x-consensus is studied for a leaderless single-integrator multi-agent system under the influence of measurement noise and communication delay,in which the upper bound of time delay can be arbitrarily large.In the second part,the p th linear x-consensus problem of headless single-integrator multi-agent systems disturbed by measurement noise and communication delay uncertainties is studied.In the third part,the mean square consensus tracking problem of leader-following double-integrators multi-agent systems disturbed by measurement noise and communication delay uncertainties is studied.The main research contents are as follows:In the first part,we assume that the time-varying communication delay is continuous and derivable with upper bound,and the upper bound of the derivative is less than constant 1.Under the condition that the communication topology contains a directed spanning tree,the asymptotically unbiased mean square linear x-consensus of a single-integrator multi-agent system affected by the measurement noise and communication delay is studied,in which the noise intensity is a function dependent on the relative state of the agent.Based on stochastic system theory,a distributed control protocol is proposed.By using the knowledge of martingale convergence,Lyapunov function theory and stochastic delay differential equations,sufficient conditions for the asymptotically unbiased mean square linear x-consensus of the system are derived.Finally,the validity of the control protocol is verified by a numerical simulation.In the second part,it is assumed that each agent can only obtain the measurement information of its own state and its neighbors,and the information obtained from its neighbors is disturbed by the measurement noise and communication delay,where the noise intensity is a function of the relative state of the agent.On the premise that the communication topology contains a directed spanning tree,a distributed control protocol is proposed.By using the knowledge of stochastic delay differential equation,martingale convergence theorem,higher order moment inequality and Lyapunov function theory,the p th linear x-consensus sufficient conditions of single-integrator multi-agent system with delay and relative-state-dependent measurement noise is derived under the directed communication topology.Finally,a simulation example is given to demonstrate the effectiveness of the control protocol.In the third part,it is assumed that each agent can only obtain the location and velocity measurement information of itself and its neighbors,and that the information obtained from neighbors is disturbed by measurement noise and communication delay,in which the intensity of noise is a function of the relative position and velocity of the agent.On the premise that the communication topology contains a directed spanning tree with the head node as the root node,a distributed control protocol is proposed.By using Halanay inequality,martingale convergence theorem,Ito isometry formula and Lyapunov function theory,the mean square tracking consensus sufficient conditions of the consistency of leader-following double-integrator multi-agent systems with relative-state-dependent measurement noise and communication delay is derived under directed communication topology.Finally,a simulation example is given to verify the effectiveness of the control protocol.