This book aims to extend existing works on consensus of multi-agent
systems systematically. The agents to be considered range from double
integrators to generic linear systems. The primary goal is to explicitly
characterize how agent parameters, which reflect both self-dynamics and
inner coupling of each agent, and switching network topologies jointly
influence the collective behaviors. A series of necessary and/or
sufficient conditions for exponential consensus are derived.
The contents of this book are as follows. Chapter 1 provides the
background and briefly reviews the advances of consensus of multi-agent
systems. Chapter 2 addresses the consensus problem of double integrators
over directed switching network topologies. It is proven that
exponential consensus can be secured under very mild conditions
incorporating the damping gain and network topology. Chapter 3 considers
generic linear systems with undirected switching network topologies.
Necessary and sufficient conditions on agent parameters and connectivity
of the communication graph for exponential consensus are provided.
Chapter 4 furthers the study of consensus for multiple generic linear
systems by considering directed switching network topologies. How agent
parameters and joint connectivity work together for reaching consensus
is characterized from an algebraic and geometric view. Chapter 5 extends
the design and analysis methodology to containment control problem,
where there exist multiple leaders. A novel analysis framework from the
perspective of state transition matrix is developed. This framework
relates containment to consensus and overcomes the difficulty of
construction of a containment error.
This book serves as a reference to the main research issues and results
on consensus of multi-agent systems. Some prerequisites for reading this
book include linear system theory, matrix theory, mathematics, and so
on.
