Consenso em sistemas multiagentes heterogêneos: uma abordagem robusta aplicada a veículos robóticos

Abstract

The word consensus, when associated with multiagent systems, is related to a control strategy that allows this group of agents through the information exchange, reaches agreement on a certain interest value. Problems such as satellite alignment, distributed sensing, page ranking, and robot formation, for example, can be modeled as a consensus problem. This has attracted more and more attention from researchers in various fields of knowledge. Considering peculiarities such as delays and heterogeneity of agents, this work presents the proposal of methods based on linear matrix inequalities, to analyze if a certain multiagent system reaches consensus. Proposed methodology for the formalization of the methods consists in rewriting the multiagent system as a linear system and, thus, analyzing the consensus through the Lyapunov-Krasovskii stabilitytheory.The efficiency of the proposed methods is verified by numerical simulations and experiments in a real problem of the formation of heterogeneous robotic vehicles.