Elsevier

Applied Mathematics and Computation

Volume 246, 1 November 2014, Pages 572-585
Applied Mathematics and Computation

Formation tracking of the second-order multi-agent systems using position-only information via impulsive control with input delays

https://doi.org/10.1016/j.amc.2014.08.059Get rights and content

Abstract

In this paper, the formation tracking problem for the second-order multi-agent system with and without input delay are investigated, respectively. The objective is to design the formation tracking algorithm such that a certain follower follows the trajectory of the leader while also maintains a certain desired geometric formation with other agents simultaneously. The impulsive algorithms are designed by using only the relative position information of the neighbors for both the cases with and without the input delay. By using properties of the Laplacian matrix and combining the stability theory of impulsive systems, necessary and sufficient conditions are derived to achieve the formation tracking of the second-order multi-agent system with and without the input delay, respectively. The numerical examples are given to illustrate the effectiveness of our theoretical results.

Introduction

In recent years, the multi-agent system (MAS) has drawn significant attention owing to its various applications in many areas, such as sensor networks [1], multiple mobile robots [2], unmanned air vehicles [3]. The research topics of MAS include consensus [4], [5], [6], rendezvous [7], [8], flocking [9], [10], formation [11], [12], [13], [14] and so on, among which formation control has been a hot topic [15], [16]. It has appealed to more and more researchers due to its broad range of applications in autonomous unmanned vehicles, navigation of spacecrafts and the cooperative control of mobile robot systems. Formation control problem in the presence of a leader which is called formation tracking problem [17] can be seen as an extension of the general formation control problem.

Over the past decades, the formation tracking problems have been studied by many researchers. In the existing literatures, many different control methods have been utilized including the sliding mode control [18], [19], adaptive control [20], [21], [22], model predictive control [23], [24], feedback control [25] and so on. The decentralized sliding mode estimators were introduced to achieve decentralized formation tracking of multiple autonomous vehicles and multiple robot system in [18], [19], respectively. The adaptive control laws were proposed to solve the formation tracking problem of electrically driven multiple mobile robots and multiple fully actuated surface vessels in [20], [21], respectively. A distributed adaptive control approach based on backstepping technique was proposed to achieve the consensus tracking and the formation control of nonholonomic mobile robots in [22]. The model predictive control was adopted to study the formation control problem for mobile robots in active target tracking problem in [23]. The model predictive control was proposed to solve the two sub-problems: trajectory control and formation control problems in [24]. The output-feedback cooperative control was applied to the formation tracking problem of unicycle-type mobile robots in [25].

The control methods mentioned above have the common feature of requiring continuous control. However, utilizing continuous control may result in much energy consumption, large amounts of communication bandwidth and short communication lifetime. Unlike the continuous control method, the impulsive control is only exerted at impulsive instants and has advantages over the above continuous control in some aspects, such as smaller control efforts, less energy consumption, less occupation of communication bandwidth and simpler implementation.

To the best of the authors’ knowledge, there are no results concerning with the formation tracking of the second-order MASs by utilizing the impulsive control with or without input delays. Therefore, the formation tracking of second-order MAS with and without input delay via impulsive control are studied in this paper. Both the directed graph and undirected graph are considered. Considering that the velocity states of agents are usually difficult to be measured in practice and the position states of agents can be obtained more easily, instead, a novel impulsive algorithm with position-only information is proposed to achieve the formation tracking for the second-order MAS. The necessary and sufficient conditions are derived for the formation tracking of the MAS with and without input delay, respectively. These results give the guidelines to design the impulsive formation tracking algorithm for the second-order MAS. The numerical examples are given to illustrate our theoretical results.

The rest of the paper is organized as follows. Some preliminaries of algebraic graph theory and matrix theory, as well as the model description and the definition of formation tracking are given in Section 2. The convergence analysis of the proposed impulsive algorithms with and without input delay under both the directed and undirected graph are formulated in Section 3. The illustrative numerical examples are given in Section 4. Some conclusions are finally drawn in Section 5.

Notations: The notations used in this paper are fairly standard. R,C and N+ denote the sets of real numbers, complex numbers and positive integers, respectively. Rn and Rm×n denote the n-dimensional Euclidean space and the space of m×n-dimensional matrices with real entries, respectively. For γC,Re(γ) and Im(γ) are the real part and the imaginary of γ, respectively. The superscript ‘T’ stands for the transpose of a matrix. The notation · refers to the vector norm. The notation denotes the Kronecker product of matrices. 1N is a column vector with all its elements equal to one, 0N is a column vector with all its elements equal to zero, IN is a N-dimensional identity matrix, diag{} refers to a diagonal matrix, or a block diagonal matrix. ρ(P) denotes the spectral radius of square matrix P.

Section snippets

Preliminaries and problem formulation

In this section, some preliminaries of algebraic graph theory and matrix theory, as well as the model descriptions and the definition of formation tracking of the system are given.

To begin with, some basic concepts and lemmas on the algebraic graph theory and the matrix theory are briefly introduced.

A directed graph G=(V,ε,A) is described with a set of nodes V=1,2,,N, a set of edges εV×V and the weighted adjacency matrix A=(aij)N×N. Node i represents the i-th agent, an ordered pair j,i

Impulsive algorithm without input delay

In this subsection, the formation tracking problem under the proposed impulsive algorithm with both directed and undirected graph are considered.

Since the output of the system are the position states, while the velocity states are unavailable, only the position information is used to design the control input ui(t), to be more specifically, only the relative positions to its neighbors are utilized to obtain the control input for each agent.

The impulsive formation tracking algorithm is designed as

Numerical simulations

In this section, some numerical examples are given to show the effectiveness of our theoretical results. Consider a second-order MAS consisting of one leader and five followers.

Assume that G is a directed graph. The Laplacian matrix L of the digraph G is given by L=20-10-1010-100-120-1-1002-1-10001. Let l=1, i.e., the follower 1 can obtain the information of the leader. Thus, B=diag1,0,0,0,0,b=1,0,0,0,0T, and the matrix Lˆ=L+B=30-10-1010-100-120-1-1002-1-10001. It is easy to see that the graph G

Conclusions

In this paper, we have investigated the formation tracking problem of the second-order MAS, where each agent can be modeled as an identical oscillator. The impulsive formation tracking algorithms with the position-only information have been proposed to achieve the formation tracking for second-order MAS for the first time. In addition, only the velocity is regulated by the control force. Moreover, the input delay has been considered in this paper. Based on the stability theory of impulsive

Acknowledgements

This work is supported by the National Natural Science Foundation of China under Grants 61374171 and 61074124, and the Program for Changjiang Scholars and Innovative Research Team in University under Grant IRT1245.

References (32)

Cited by (28)

  • Multi-group formation tracking control via impulsive strategy

    2020, Neurocomputing
    Citation Excerpt :

    Therefore, it is natural to further study the impulsive formation control with time-delay. For delayed impulsive systems, there are lots of results according to different types of delay, such as constant delay [17] or time-varying delay [18–21], bounded [28–30] or unbounded delay [24], distributed delay [25] and stochastic delay [31,32]. There are many recent attempts to investigate impulsive consensus problems of MASs with time-delay, seen in [33–35].

  • Relay cooperative tracking control of networked nonlinear multi-agent systems with communication delays: A data-driven method

    2019, Neurocomputing
    Citation Excerpt :

    Generally, there are two kinds of delays: one is the delay that the agent itself has, called time-delay [17], and the other is the delay brought by the communication process, called communication delay [18]. Paper [19,20] were researched the formation tracking problems with delays. Unlike most studies on tracking in time domain, paper [21] studied the consensus tracking problem of MASs with input delays from frequency domain perspective.

  • Static consensus of second-order multi-agent systems with impulsive algorithm and time-delays

    2017, Neurocomputing
    Citation Excerpt :

    For the case of aperiodic sampling and switching topology, [21] proved the impulsive consensus of second-order agents where only position measurements were available, by using the property of stochastic matrices. More results on consensus via impulsive control can be found in [27–30], and the references therein. Moreover, the disturbance of communication time-delays is unavoidable in real networks due to limited communication capacity of sensing or transmitting equipments.

View all citing articles on Scopus
View full text