dc.description.abstract | A multi-agent system (MAS) is a dynamic system that consists of a group of interacting agents
distributed over a network. In the past decades, the study of distributed coordination of multi-agent
systems has been widely attracted by many groups of researchers such as mathematicians,
engineers, physicists, and others. This is partly due to various applications in many areas, including spacecraft formation
flying, multiple robot coordination, flocking, consensus or
synchronization, cooperative control of vehicle formations, etc. As one of the most important
problems in distributed coordination, consensus means that a group of agents achieves
an agreement on a common value by designing the control law which is based on the information
received by interacting with neighbors. There are many consensus methods that have been
studied in recent years. Some problems focused on seeking the consensus of continuous-time (CT)
multi-agent systems or discrete-time (DT) multi-agent systems, the others considered consensus problems
on hybrid systems which are dynamical systems involving the interaction of continuous and
discrete dynamics. Most consensus algorithms have been proposed for the multi-agent systems,
but most results of consensus analysis are on the situation that all agents are continuous-time
or discrete-time dynamic behavior.
There are, however, some practical problems that the discrete-time and continuous-time dynamic agents coexist
and interact with each other at the same time. Thus, it is reasonable to study consensus problems
in such hybrid multi-agent systems (HMASs).
Generally, the consensus protocols are designed to ensure that the states of all agents converge to a common value. However, up to date, in many practical problems, the states of agents may converge to prescribed ratios rather than a common value, such as compartmental mass-action systems, water distribution systems, and multiscale coordination control between spacecrafts and their simulating vehicles on ground. To deal with this problem, the scaled consensus problem has been introduced, where all
agents will converge to the assigned proportions. Different from the standard consensus, where a group of agents seek to agree on a common quantity depending on the states of agents, scaled consensus implies that the state of each agent will approach prescribed ratios in the asymptote.
So this work aims to study the (scaled) consensus problems in hybrid multi-agent systems under fixed and switching topologies including linear and nonlinear dynamics. Furthermore, we study consensus problems with communication delays, external perturbations, finite-time (scaled) consensus problems and also apply to the random networks.- | en |