Asynchronous Distributed Algorithms for Heading Consensus of Multi‐Agent Systems with Communication Delay

This paper investigates the flocking problem of multi‐agents following a leader with communication delays in a noisy environment. Based on potential fields and the LaSalle‐type theorem for stochastic differential delay equations, by introducing the adaptive protocol compensating for the desired velocity, a new neighbor‐based flocking protocol is proposed such that all the agents move with a virtual leader's velocity almost surely, and avoidance of collision between the agents is ensured. A numerical example is given to illustrate the effectiveness of the proposed methods.

Section: Introductionmentioning

confidence: 99%

Flocking of Multi‐Agents Following a Leader with Adaptive Protocol in a Noisy Environment

Li¹,

Liu²,

Zhang³

2013

“…In the last decade, the consensus problem has been intensively investigated in various disciplines, such as mathematics, biology, physics, computer science and control engineering. Most of the works focus on the study of system stability under different cases, fixed and switched topology, time delay, and quantization communication, see [1][2][3][4][5][6][7][8][9][10][11]. Much effort has also been directed toward to the leader-follower system, where the leaders as an intervention push all the followers reaching consensus on the leaders' state; see Wen et al [12], Chen et al [13], Su et al [14,15], Lu et al [16].…”

Section: Introductionmentioning

confidence: 99%

Optimal Controlled Nodes Selection for Fast Consensus

Yang¹,

Wang²,

Wang³

et al. 2015

Motivated by the significance of optimizing network performance with limited resources in real applications, we investigate the problem of achieving fast and desired consensus in a complex network via pinning control a fraction of nodes in the network. An optimization problem is proposed to select p optimal pinned nodes to guarantee fastest consensus, which is relaxed to a Mixed-Integer Semi-Definite Program. In the single pinned node case (i.e., p = 1), we find that the optimal pinned node is the node with highest min-max distance centrality within the nodes with highest closeness centrality in the network. In the multiple pinned nodes case (i.e., p > 1), we find that betweenness centrality could be used as a suboptimal index to select pinned nodes.

“…In [8], the authors surveyed some recent investigations on collective group behaviors, and they designed three predictive mechanisms, both centralized and decentralized, for flocks with or without leaders. In [9], two asynchronous distributed protocols for the heading consensus of a multi-agent group were proposed. Due to the fact that, in practical applications, multiagent systems usually are affected by disturbances, such as parameter uncertainties, communication noises, and unknown nonlinear dynamics, the effectiveness of these disturbances and how to design a controller to overcome these kinds of disturbances should be considered.…”

Section: Introductionmentioning

confidence: 99%

Distributed Robust H_∞ Consensus for Multi‐Agent Systems with Nonlinear Dynamics and Parameter Uncertainties

Wang¹,

Wu²

2014

In this study, the consensus problem is discussed for a class of multi-agent systems with unknown Lipschitz nonlinear dynamics, external disturbances, and parameter uncertainties under a fixed directed graph. By some model transformations, the consensus problem becomes a reduced-order H∞ control problem. Based on the reduced-order control problem, sufficient conditions to achieve the consensus with the desired H∞ performance are presented. These conditions are to check the solvability of only one linear matrix inequality (LMI) and the dimension of the LMI independent of the number of agents in the network. Moreover, an algorithm is proposed to design the dynamic output feedback controller. Simulation on networked multi-agents is provided to show the effectiveness of the theoretical results. Fig. 3. x2(t) error trajectories of all agents.Fig. 4. Energy trajectories of the measure function z and disturbances w1 and w2.