Observer-based consensus control strategy for multi-agent system with communication time delay

Abstract

Abstract-In this paper, we discuss a target-enclosing problem for a group of multiple nonholonomic agents in a plane. The proposed strategies guarantee that multiple agents' coordination finally results in a circular formation enclosing the targetobject which moves in the plane. Firstly, virtual agents for the feedback linearization of the real nonholonomic agents are introduced. Secondly, we propose the target-enclosing control laws based on the consensus algorithm to the virtual agents. Algebraic graph theory and consensus algorithm are employed to prove convergence and stability of the enclosing problem. Finally, experiments are provided to demonstrate the effectiveness of the proposed control laws.

Experimental validation of cooperative formation control with collision avoidance for a multi-UAV system

“…A major strategy for formation control is to apply a consensus algorithm [2]- [5]. We have been studying cooperative control problems for multi-vehicle systems [6]- [9] with particular focus on solving formation control problems using a consensus algorithm. We expressed the UAV dynamics on the horizontal plane as a fourth-order system and the dynamics in the vertical direction as a second-order system separately, then proposed two consensus-based control algorithms for a multi-UAV system to cooperatively fly in formation on the horizontal plane and in the vertical direction [8], [9].…”

Formation Control of UAVs with a Fourth-Order Flight Dynamics

“…The authors have been studying cooperative control problems with a multi-vehicle system [17]- [21], especially focusing on formation control problems using a consensus algorithm. In [19], The authors proposed a control algorithm for multiple columnar vehicles to surround a moving columnar object cooperatively without depending on a network structure.…”

“…The results in a first-order system are directly applied to most works of a second-order system. Works of a linear system [3], [9], [13]- [15] and [17], however, are not directly extended from the results in a first-order system, because it is difficult to extend the algorithms for a first-order system directly to the problems with a linear system. Therefore, most authors who consider a linear system mainly use complicated ways such as an optimization approach and a LMI (Linear Matrix Inequality).…”

Virtual structure based target-enclosing strategies for nonholonomic agents

“…On the other hand, consensus algorithm based formation control strategies for multi-agent systems are proposed in [14]- [17]. Ren [15] proposed the formation control strategies for multi-agent systems where the information states of each agent approach a common time-varying reference state.…”