Discrete-Time Dynamic Consensus on the Max Value

In this paper, we propose a novel consensus protocol for discrete-time multi-agent systems (MAS), which solves the dynamic consensus problem on the max value, the so-called dynamic max-consensus problem. In the dynamic max consensus problem, the objective of each agent is to estimate the time-varying value of the maximum instantaneous value among the reference signals associated to the agents in the network, by exploiting only local interactions. The proposed interac tion protocol enables the agents to solve this problem with an a priori bounded error, without exchange of input information among the agents. Furthermore, the proposed protocol can be tuned by means of a tuning parameter, enabling a trade-off between convergence time and steady-state error. We also provide a preliminary characteriza tion of the maximum relative tracking error. Numerical simulations corroborate the theoretical analysis of the convergence properties of the proposed protocol.