Scaled leader-following consensus of networked agents under packet losses

This paper is concerned with the scaled leader-following consensus in discrete-time multi-agent systems with packet losses. The scaled leader-following consensus means that agents eventually reach convergence at their own scale following the state of the leader. In comparison to existing scaled consensus literature, the environment of packet losses is introduced into scaled consensus issue for the first time in this article, in which the success or failure of information transmission in each link is independently represented by a Bernoulli random variable. Based on a stochastic distributed control protocol, it is shown that the scaled leader-following consensus with packet losses can be achieved if and only if the interaction topology contains a spanning tree rooted at the leader. Finally, a simulation example is shown to illustrate the effectiveness and feasibility of the theoretical result.